MIT’s Breakthrough Formula for Startup Success with Gene Keselman
38m 14s
The conversation between Ryan Connell and Gene Kesselman focuses on the dual-use readiness model developed at MIT for defense innovation. Kesselman explains that beyond technology readiness levels (TRL), startups must also consider customer and funding readiness in both mission (defense) and commercial contexts. The model provides a common lexicon to bridge the gap between startups and government entities, which often struggle to understand each other. It helps startups assess their current state and identify next steps, such as pursuing phase one grants or MOUs. Kesselman highlights that the model is open-source and used by programs like NATO Diana, with an AI assistant being added to answer queries and demystify processes. Regarding strategy, he notes no proven advantage to starting commercial-first versus defense-first; instead, startups should remain opportunistic, pivoting based on funding opportunities like DARPA grants. The dual-use approach is gaining acceptance, as VCs now recognize the value of government funding and interesting problems, reducing previous stigmas. Overall, the model serves as a roadmap for startups navigating the complex defense innovation ecosystem, emphasizing communication and adaptability over rigid planning.
What we realize is beyond technology, there's two other really main focuses for startups. There's customer and funding. If we can capture your readiness on where you are in finding and capturing customer, finding and capturing funding and developing your technology, I think you can really well define a company at that point. This is as close to a map as I think we have. We choose to go to the moon and this decay and do the other things, not because they are easy, but because they are hard. Through our God and your bonds, we price the Germans before he got it. You and I have a rendezvous with Destiny. Hey, this is Ryan Connell with the Chief Digital and Artificial Intelligence Office. Join here today with Gene Kesselman. Gene, how you doing? I'm great. Thank you, Ryan, for having me. Yeah, looking forward to diving in today. I know you were over at MIT. I believe director of mission innovation experimentation. Love to talk to a fellow person that hasn't ex at the end of their title. You want to give a quick background in introduction? Sure. I am at MIT. That's my full time day job. I run a couple centers there. What is MIX? That's focuses on dual use and defense tech, national security innovation. All those things have fallen to that kind of rubric. I run our internal venture studio at MIT called protoventures where we work with MIT research and try to spin out new market companies out of MIT. I'm also a lecturer at Sloan at the Sloan School. I teach mostly around dual use and defense innovation. My part on job is still with the Air Force. I'm a kernel in the reserves. I work at the Pentagon for the Air Force. I've been in about 12 years in active duty and remaining 12 on reserves. That's me in short. Awesome. You want to talk a little bit about your role at MIT. In the X role, are you in a lecture type capacity or are you running a program? What does that look like? Both. Yeah. All the above. I started off with literally just a huge demand signal when I got first got to MIT from the defense folks from defense industry about just wanting to work with us on innovation. My role when I first got to MIT was the director of the Innovation Initiative, which was focused completely internally at MIT. We were looking at all these interventions and programs and ecosystem around innovation inside MIT. Just by the nature of my background and having spent most of my adult life in the defense world and DOD, all of this stuff followed me up to MIT. We had nothing. We were not focused at all on national security and global security. All of these were still used and everything. But we kept getting more and more people coming say, we really want to work at MIT. We want to learn how you guys do innovation and things like that. So we stood up a program. I hired a program manager just to respond to the demand signal of that. Over the years, as we've gotten more programs running, it's turned into its own center. We have programs, research, teaching, all sorts of stuff that goes along with having a center. That's all really just completely come up from all demand. Never thought that this would be what we'd be doing at MIT eight years ago. So just help me orient, right? Let's say I'm interested in defense innovation. I certainly am. Is this something that organizationally I hire MIT to support? Or am I going in and saying I want to sign up academically for your course, Gene? Or how does someone get involved? Sure. So yeah, so we're a university, right? So a lot of times this comes up as like what can you do with the university? Well, we have two like explicit clear pillars of every university does. It's teaching and research, right? So we at MIT do both. Some universities don't do a lot of research, but we do a lot of both. So yes, we have a dual use class that we teach actually this week. We're going to have a class open, open enrollment class for dual use when we teach kind of the basics one on one of a dual use strategy. We also teach that class to NATO for NATO, through the NATO Diana program. We also have a class with special operations command for MIT and and and students around Cambridge or our design build class for so-com problems, right? So teaching we have a good amount of stuff on teaching. We have a bunch of research programs where we partner with startups on STTR, phase one of phase two grants to do kind of dual use introduction and then real research on the startups. And you know, we have a bunch of other things around translation where we help with translating either ideas into startups like the venture studio. So really it's kind of that bucket of stuff. We don't really like if you're a company that really depends on where you fall or kind of on the spectrum of dual use. So we kind of almost work mostly with startups or with government organizations. So so-com, NATO, a few Air Force or startup. Like if you're a big system integrator and you want to work with MIT, that's basically funded research, right? You're going to be funding different off different research centers around MIT. That's not me. We're kind of agnostic of any specific research area. You got it. No, that's helpful. You mentioned dual use a couple of times and I know that's, you brought up a readiness model we talked about right before we get started here. You also use the term earlier in already Rubric. So I got the impression that you're someone that likes to have a model or a Rubric in place. So why don't I just let you talk about the dual use readiness model and I'll have some questions I'm sure from there. So that model was created for our work with NATO Diana because we were going to be working with all these startups. So we needed to evaluate kind of have some kind of model where we evaluate startups on now, not just what we obviously in our community all know the TRL scale, right? Like that is a a very easy heuristic for understanding kind of approximately where a company stands and technology development. So if I told you I'm a TRL six company, you would know almost within a bounds exactly what kind of company I am right where I fall on your expectations. As dual use and kind of this defense tech world started becoming more more interesting, more prevalent around the globe. Like we realized there's not a common language. There's not a common Rubric. There's not a common model for people to understand that if I said I'm a CFRL seven company, you would know what CFRLs, but you wouldn't you don't understand what that is. So one of the things we set out to do is not just do these programs, but also do thought leadership and language and common lexicon around the stuff because it doesn't exist, right? That's one of the first, if not biggest complaint from both sides of this marketplace. So from the startups and the government, folks is we don't understand each other, right? The startups, they don't know how to access. It's just opaque, you know, non-pores membrane between them and the government. And likewise, I think some some government folks you probably can relate to think they understand how the other half, but they just don't. They don't understand what it's like to be in a company in a startup, like working, you know, in a different field than the government. So the, I think the most important thing with that, we talked about that article that we published recently and the model and all of these things is not just to drop, you know, a fancy model and someone's quite say, here you go, like use this. It's to accompany a language, you know, lexicon for everybody to have a joint definition of things. And so we all can be speaking the same language when we're talking to each other. And I think that's by far the biggest gap in our industry right now is that just people don't understand the other side. Yeah. So all right. So I think the TRL, the technology readiness level is a good, a good level set of like the way you kind of frame that. So what is the model measuring? Is it technology or is it like readiness or is it, is it just like maybe a startup's ability to communicate government or a government's ability to communicate like I'm just trying to understand what the model is. Sure. So you could think of it as like three vectors and two dimensions, right? So it's way more complicated than it really is, but I'm just giving you kind of, so what we realize is beyond technology, there's two other really main focuses for startups, right? It's customer and funding. Okay. So if we can capture your readiness when you can quotes for those that aren't watching on the where you are kind of in finding and capturing customer, finding and capturing funding and developing your technology, I think you can really well define a company at that point. Right. I think you can very like closely match that with other companies in like in the like readiness stages, right? And then kind of so those are the three the three vectors and then the the two dimensions are.
mission and commercial. Right? So they're not the same thing. Right? We don't need to treat them like we can cross-pair like that. They're not, you can't equate them, you know, directly. They're very unique in many different ways, but in both of those cases, you are still going to go after funding and customers in each of those, as well as developing technology, sometimes the same technology, sometimes it's different, sometimes what you're developing as a product or technology for a commercial side, you have to adapt for the mission. Right? So I think if you think like that, and again, that's going to be a more complicated version, but basically what we're capturing is mission customer and mission funding, commercial customer, commercial funding, and technology readiness. And on those five levels, if I tell you, if you tell me as a startup where you are in readiness level on all five of those, I have a very good idea of what kind of company you are. And then I think the most important thing, if I may just like land this plane, is not even where you are, but where like the next steps are. Right? So you can look at this guessing, I'm a three, I don't know anything about the mission, I've never worked with a DOD, I know I'm a one or two, right? But then you kind of can look at what three, four, five is and you get an idea of where you want to go. Right? Oh, so I need to go after a phase one, I need to get a MOU, I get, you know, things like that. And you know, it's like a two to us inside. This might seem like well, of course, I guess that's what everybody has to do. But if you've never dealt with this, you have no idea that this is the process, then you need a roadmap summary, you need a map. That's super interesting. Yeah, I'm just thinking like how many, if you're, if you can say like, how many companies have you kind of put through or put through this process? So lots informally like through, just letting people use the model, like we have an accelerator and can we use your model, of course, like this, you know, it's open source. We're not charging for just use the model. All I ask for is the data back. Formally, again, NATO Diana used it for the first cohort of startups that came through last year. And so the new cohort is starting, well, it already started, but I'll be going out there to teach next week from that cohort. So it's certainly one of those things where we get more and more companies in, we're going to get better feedback. But it also isn't that complicated, right? Like there's a lot to tweak, I'm sure over the years, but really what this is is again, it's to help them. It's not for them to, I don't need a huge data set and to find out that we had five, five thousand companies that were commercial funding writing the level six, right? I don't know that that tells me anything. Well, that's where I'm like that's where I want to go with this, right? Like I'm just a data nerd sometimes on the inside. So I'm like, gosh, is there a secret code to our VCs out there? Like if it's a 1379, it's, you know, it's going to be successful 90% of the time or whatever that secret code is, you know, I wish, because then I would just lock it down and then you know, sell it for millions. No, I don't think it works that way. I don't think that's that's the way it would be used. I think it is much more of a communication and, you know, getting a foothold on the strategy as much as it is. I don't think it's a data trod, you know, try to travel data. I will say though in terms of AI integration, it is the place where we are first starting to integrate AI from a product standpoint into this model. And that's and I think because probably you can relate to this, it's the place where the greatest amount of value I think can be attained from somebody coming in and querying the model and getting explanations that are that come from, you know, these large language models. So just to give you an example of how we're looking at this, I started playing around a while back with just, you know, some open AI chat GPT kind of, you know, in the three, four time frame of training it, not training the model, but training, you know, giving it a data to understand this model or to understand the CFRL, see all these different writing levels. And it turns out it really understands them very well. It inherent to the models that are trained on, you know, trillions of tokens on every piece of information out there inherent in those models is an understanding of how basic entrepreneurship works, right? So basic TRL works, how companies progress through commercial readiness just without it being explicitly talked about in a scale, but just like how startups work. And so when I add kind of this mission focus, there's enough understanding of this the parallel side of defense within these models and kind of acquisitions and all this other other stuff, non-delutive funding. It's already, it's already been trained in, like they're not purpose built for that. But I was afraid that there wouldn't be enough context in the models to understand to be able, but if I just fed them the models and gave them some of our thought leadership writing stuff, that it wouldn't, would not be able to make sense of it. But it does. And so what we're doing is we're actually on our website that's built for this model. We're actually going to be adding an AI assistant where you can not just look at what number you are, but you can ask you questions. Like why do I have to do a phase one? Like what does a phase one even mean? Like, okay, now tell me more about what commercial, you know, a B series B round is, if I don't have any idea what that is. And I think that would be another huge step into breaking down this language divide. And I think that's a, you know, my favorite use of the AI tour right now. Awesome. No, that sounds great. So obviously, you know, you kind of answered the last question in terms of like using that data as any sort of predictor. But maybe I'll zoom out one level, but but similar question. With the dual use, do you find that there's any more success in a company that starts off commercial and then transitions to DOD versus a company that starts out with the DOD focus and then finds a commercial customer? There's no empirical proof that either one is is better. I think the only evidence we have right now is that the only way to think about it is opportunistically. And if you, if you kind of read anything we wrote, that's basically the entire premise is that rather than going in with assumptions about what you're going to be doing, you need to like literally treat every day as an opportunistic way to find customers and funding. And if tomorrow, you got a DARPA grant for $2 million and you want to go, you have to go pivot real quick and focus on, you know, delivering that kind of capability, you do it. And then after that's done, if it's a dead end, you go back to focusing on commercial and you know what, maybe you never come back to DOD. It's a defense mission, right? Maybe after that experience, you're so chated. And so like, you know, never again, I don't want to do this again. But you got $2 million and you know, some runway to develop your technology, right? And now you're like, you know what, back to commercial, really focus on that. Let's go build upon it. Let's go sell it. Let's get to the consumer. I think that is a perfectly good strategy. And I think the one thing we know is that I can't take that and port it over to you in your startup and say, just do this. Because everything's different, too many different variables. But if you do learn something, it's that it takes a lot of, you know, like a VC situation where like, you know, if you think you're going to pick up the phone the first day and call a VC and get a, you know, a safe from that conversation, you know, you obviously haven't done it. You got to do it 50, 100 times before anyone's going to, you know, take a lead on around or anything like that. Well, the same thing on the DOD side, you got to, you got to, you got to turn, you got to do a lot of applications, grant applications, a lot of calls and customer discovery. And so I think the only thing it shows you is that through that process, you might want, you might look at another startup and say, I want to do that. I want to go get that DARPA brand. Okay. We'll go apply for that DARPA brand. Well, you didn't get it. They did. So you have some choices. You can apply for 10 more or you can just say, I quit and I'm going to, right now I'm going to focus on the other side. That's, I think that's what you can learn from other examples like that. Yeah. So how do you like, so just there is, obviously you can give out advice as part of the accelerator and all that, but like kind of navigating what you just described, it seems obvious that, you know, for someone that's a founder, they kind of, you know, creates their business plan. You have a plan that probably looks like a straight line. And then you got a customer that's kind of like, well, can you do this and pulls you one direction? And as you're working that project or maybe down, you know, coming to a downturn that project, someone on this side's kind of pulling you over here. And it just seems so easy to get distracted from, I'd say, maybe what your original business plan was. And how do you navigate that? Well, I mean, that's, that's, that's a really interesting question. And one we get a lot is, is, well, you know, it used to be that VCs, anyone on the risk capital kind of commercial side would be merely turned off if you were pursuing government grants because it was just too, too distracting, too complicated, takes you off the path, takes you off the kind of your focus, you're too distracted. I think two things factor and say one is, I don't think that gives people enough credit. Like, startup founders, they can walk and chew gum at the same time. I've seen very
very capable people think that they can't, you know, like how could I possibly go write these, you know, NIH grants and do, well, you can. I trust me. I've seen lots of people do, you can't do it. You know, it's a lot of work, but you can't do it. And once you get really good at it, it becomes much easier. I think there's a stigma attached to it. And I think there was that belief that you couldn't do both. I think both those things are gone. I mean, I think if there's one really kind of key factor into why dual use is having at some moment right now, is I think that a lot of those kind of predispositions, those those, you know, wives' tales and those beliefs that, you know, this is a bad mark and you can't do it are gone. I think there's just enough examples now. And everybody sees this huge pool of money over here and it's being underutilized and desperately needed and has some of the most interesting problems. Sure. And now the industry is saying, why not? Like what's wrong with this? And when more and more dual use VC stand up, you know, they have the big VC starting their own kind of dual use funds. And why combinator has a call for startups for defense tech, like, you know, the culture has changed. And it's not just that startups are better. The culture has changed such that this makes, this is really having a moment for dual use. Yeah. And like so, you know, I'll share a little bit anecdotal, but based on a real story where, you know, I know someone that was starting up focused on DOD, right? It was AI type solution, human type interaction or at least intended to replicate human behaviors. And focused on DOD, they went through the Siversiter program, Siversiter effectively forced them to create a whole commercialization plan. And you know, that wasn't part of the founders vision, right? It was like, okay, now I got to go find a commercial use case and non-governmental use case for this technology. All while, you know, you got investors from foreign investors that are trying to buy his company and he's trying to be a patriot and like stay with working with DOD. We're not easy to work with. And in the long story short, you got, you know, effectively pulled so many ways and so thinned out that he closed the business. And just like I couldn't do commercial and DOD at the same time, but we have kind of a structure that promotes that with the Siversiter program. So I don't know if you've experienced something like that, but that's a story that's near and dear to me. I think in most startups, I know we love to have that problem. Okay. To be honest with you, like, yeah, that's not the ending I expected. I thought he was here. She was going to be like, you know, super successful one side or the other. I mean, if you have all of this demand and you have people wanting to buy you, you know, what's the point of a startup? The point of a startup is to create, you know, economic prosperity for you and other people and to also deliver a capability, right? To either, you know, a product or to the mission. So, you know, if you do it, you should consider yourself successful. I mean, I think shattering down is not the option I've seen most of the time. It's usually what happens is you have to make a really hard decision and you never know if it's the right decision and it's really painstaking. It's really, really difficult and, you know, really sends a lot of these founders into, you know, self-doubt. Am I making the right decision? Am I wasting too much time? Am I going after this and, you know, banging in my head against this wall for nothing? But usually it works out. If they have that kind of demand that doesn't kind of matter which side they go with because it's going to be okay because you're going to keep generating funding and revenue. Yeah. That makes sense. Awesome. I know you referenced it earlier in the conversation, but you recently wrote an article, I think you wrote an article for War on the Rocks. I don't know if you wanted to, I don't know if it's everything you've just described because you added it to my to-do list to read right before we got on. So I'll let you hit on that. If there's anything else that was in that article, you haven't hit on already. No, it's really, and it's everything I just said is basically the whole, the whole article's premise is that the do-use is a strategy and how to category and that. So if you think about it as a strategy, you have to, you have to use every day and think about every day and you have to make opportunities to the decisions and you have a language that's common and common with both sides that you want to work with, then you'll be okay. Then, you know, you will make, you know, the right decisions as a startup founder. If you think like I'm just stuck in a do-use and I'm a defense company and I will always be defense company, it's the only thing I can be. Then, you know, that's, it can be successful that way. Of course, we know lots of companies that have focused only on defense have been okay. But I think, predominantly, even the government will say, I'm sure you will say this, we want you to become a commercially successful company. To scale, to add to the industrial base, you need to become a commercially successful company. The government on its own is not a good customer. But the government buying things that it needs, capabilities and needs from commercial companies is a really successful model. Yeah. I mean, yeah, it makes sense, right? Like if I have a down budget year and I can't afford to give you as much money this year, I don't want you to go to a business. Like I want you to have other forms of revenue coming in completely. Yeah. And then you have, you know, all kinds of factors. You can't even anticipate like a new administration comes in and they just halt all contracting. Like they did today. Yeah. Like, okay, like you're waiting for that OTA contract to come in today and all of a sudden you get to call and the government's decided they're going to just halt contracting for a few weeks. Okay. Well, I got to make payroll. So you got to have that commercial side. Yeah. No make sense. On the commercial side, I'm curious, I came up actually in a conversation I was someone earlier this week, but do you envision the commercial side to mean like non-governmental related or do you envision like, so if it's like GovCon, for example, like there are two sides of that. Like I could sell a product to the government that would be for the government or I could sell a product to defense contractors that are working for the like that are sure that would use this product. Is that commercialization or are you trying to get out of defense? It's out of fuzzy line. I think it's more discreet than a fuzzy line, but in some ways it's basically dilute versus not dilute of funding. It's kind of a good way to think about this. Certainly like the goal is to get to revenue and they go see it to revenue from somewhere. And but that takes a lot of time, especially in the government to get to a point where you're selling items at scale is a very long time. But the funding side of it, you can think about it as non-dilutive and dilutive where you can continue to grow your company, you grow your technology through non-dilutive funding, through grants and all sorts of programs with the government. And then grow your company through private capital and VC and all this sort of stuff. And yes, in between there is potentially having a intermediary like a large system integrator. In the US, you have a lot more options. As I mentioned, they won't go to NATO a lot. Outside of the US and almost everywhere else, you have to have an, you can't go directed to the government. They don't have a SBIR equivalent. They don't have a really explicit kind of phased approach to getting to that competitive contract. So you have to go through a large system integrator. Basically, if you have to have the government go to someone else and say, we in Poland, we really want this. You, company that we have, you know, a huge IDIQ-esque thing, go get that. And then you're dealing with it in our intermediary. They're going to be buying a product, but they're not going to be funding you unless you do like a joint research thing. So that's really the, one of the key kind of distinctions between that is, is understanding that what we're talking about is usually early stage funding, not revenue. And so really kind of the distinction there's delude of a non-deludive. Sure. Okay, no, that makes sense. That makes sense. All right. Yeah. I wanted to pick your brain too on the, I think you talked venture studios, protoventures. I think that's what you called it. So this is an accelerator program that you all have within MIT. So it's actually a venture studio. It's taking a commercial venture studio model and putting it inside MIT. Got it. It's, help me understand that a little bit more. Is it, are you matchmaking with VCs effectively? No. No. I think we'd love to figure out ways for VCs are very interested, but we're very unique. We don't take any equity. We don't take any carry off of any investment that MIT is putting into the startup. So the way it works, in fact, I just kind of do the VC, the venture studio kind of 101 real quick. The commercially venture studio is basically taking LP money or investor money and instead of investing into discrete startups that come along to match your thesis and have certain criteria. You're saying we're going to build those startups internally, right? We're going to, we're going to bring really, really smart people. We're going to teach them how to kind of surround them with support, teach them how to do the venture building, the EIR sort of process. And then anything that comes out, the venture studio is going to have a large stake in any of these startups. The model that we actually built this on is, if you've ever heard of a studio called Flagship Pioneering. It's one of the, you know, in the top five of all VC kind of funds in the country, but it's a venture studio where all of their LP money goes internal. Like they just build the biotech startup.
and in fact, they're the ones that build Moderna. And so that's the model that we looked at is like, why can't we do this? What better sandbox to build really impactful startups than MIT? But it doesn't transfer one for one because we get where university, we can't do things like take equity and startups coming out of research here. That's for technology licensing and for the MIT corporation. So the one thing that we did is we were able to track super, super capable, I call them unicorn human builders, which are these like incredible technologists, PhDs that are also entrepreneurs. And we basically surround them with MIT and release them into the ecosystem for two years. They embed in the research environment. They embed in these labs. They get to know everybody, get to know everything that's going on in whatever the studio is on. So right now our studios and fusion and clean energy. And so by the end of that first year, they've discovered everything that's going on at MIT, fusion and clean energy. And then they start building teams and startups out of the most promising research. It's really kind of a translational process. And it's like, it's learning how to do that thing that lots and lots of corporations are trying to do. Lots of research organizations, universities are trying to do is create a process around translation rather than it being completely synodipitous. It happens a lot, like startups spin out of universities, but it happens only out of synodipity. Like all of these variables have to align. And what we're trying to do is we're creating, by using a studio model to make sure that those variables align or just reduce the number of variables. So that if you are launching a startup at the end of two years, you're almost assured or highly likely that it will be successful because all these variables have been taken out of the equation. So you're-- I don't want to say purpose, but your angle for doing this all is research based to align with figuring out what the best process is to reduce number of times that startups aren't spun out. Sure. You can look at it exactly like that. You could think about it like there's our normal distribution of translation that happens to university, right? There's lots of good amount of PIs, especially the place like MIT professors and researchers, that are really good at it and have spun out multiple startups out of their research, right? Because they really know the ingredients to making a work, a good PhD or postdoc or a team of PhDs and postdocs. A good idea, understanding how market analysis works, all these sort of things, having the real drive to do it rather than just sitting back and doing research. And then there's the far end of the normal distribution, the third segment on the other side, which is the theoretical mathematicians that if you went into their office and said, hey, I want to do a startup with you. They'd be like, I'm here to do math. I should be able to-- Yeah, yeah, yeah. Stop. I don't like that word startup. And so in between, there's this really broad range of researchers. And this is the same for every research university. And mostly every research organization in the world is that there's this broad range of people that are either don't know anything about translation but don't realize that they're sitting on something that could possibly be a really interesting company. Don't know where to start, right? They think that this is something they have no idea where to their scientists or engineer. They've never done this business stuff. Or if it gets too hard or a lot of other reasons, the opportunity costs are just too much. I'm not giving up my tenure to start a company. But if somebody else came along, I would be happy to participate. And so really, what Protoventures is or what Adventure Studio inside of university is to serve that community in the middle of that distribution and say, we can help you. We'll either build some teams to help you. We'll work directly with you to start the startup. Or I will be the CEO of this new startup and you will be on the board and you'll get an equity stake. - Sure, got it. It's the success is predicated on just two variables at that point. How good your venture builders are and how good the research is or how good you are finding good research that could be commercializable. - No, that makes sense. - Okay, awesome. No, that helps. I appreciate the follow up. So getting close to time here, I want to give you an opportunity to be king for a day. I'd like to do this with a lot of the guests that come on. But any big thing that runs to your mind in terms of potential changes or things you would do if you were king for a day within this entire ecosystem that we're talking about. - I teach, as I mentioned, I teach the NATO data cohort, I do the intro to dual use in defense ecosystems. And I've done a lot of research. I kind of a European market and all these other different markets. And I think I know that we don't realize what an enormous head start and enormous advantage we have on every other country in the world when it comes to this sort of stuff. We are the 900-pongarillo. We know that, we have this enormous budget, huge percentage of our GDP goes towards defense. It's in our culture. But I don't think people here really understand how different that is than the rest of the world. Aside from Israel and unfortunately Ukraine right now, like there's just not a lot of places that are understant, but it's happening, the changes happening quickly, especially those countries on the border of Ukraine. They're figuring out very quickly though. This defense thing, I think there's something too. Maybe we should take this more seriously. So I think if I were king for a day, I would rather than maybe change anything in the US. And I think because we do have such a head start, and I know we have a million problems, I'm not saying we don't have problems. We certainly can change a lot of different things. But there's lots of people publishing all of the things that we could do for our acquisition process and everything else right now. So I'm not gonna, that ground has been tried in this being the rest of the world. Yeah. I think what I would look at is the rest of the world, because I think it's really, really important for the rest of the world to catch up and catch up quick, for the sake of our, you know, combined our collective global security. There's just a little bit too much instability right now. So I think if we could really, at the highest diplomat and economic policy maker levels, really look at the models that we have here, like the SBIR, STTR program, the OTAs, all of these different vehicles that allow us to have this huge advantage, and really bring it over to our partners. Sure. And really, like in official capacities, and in real effort to be like, these are doable. You could pass a law that says every agency over $100 million on D budget has to have 3.2% of its budget to go towards the small business program. Yeah. We did it, we did it four years ago. Like you could do that tomorrow. And I think that's what I would, if I would go be king, not like a van, I'm gonna be king evangelist. (laughing) And I would really try to get everybody, not our partners, not our heaven's sake. I would leave them blind, but like, here's a few models, use them, you will be better off and you will build your ecosystem. No, I like that. Awesome. Hey, Gene, I really appreciate you coming on today. This is a great episode. Pleasure, my name is Gene. Yeah, thank you so much. Thank you. Take care. (upbeat music) (upbeat music)
Podcast Summary
Key Points:
A dual-use readiness model is introduced, measuring startups across three vectors: technology, customer, and funding, with two dimensions (mission and commercial).
The model aims to create a common language between startups and government, addressing the communication gap in defense innovation.
An AI assistant is being integrated to help startups navigate the model, providing explanations and guidance on next steps.
There is no empirical evidence that starting commercial-first or defense-first is superior; success depends on opportunistic adaptation.
The stigma against dual-use startups (pursuing both commercial and government funding) is fading, with more examples and support from VCs.
Summary:
The conversation between Ryan Connell and Gene Kesselman focuses on the dual-use readiness model developed at MIT for defense innovation. Kesselman explains that beyond technology readiness levels (TRL), startups must also consider customer and funding readiness in both mission (defense) and commercial contexts. The model provides a common lexicon to bridge the gap between startups and government entities, which often struggle to understand each other.
It helps startups assess their current state and identify next steps, such as pursuing phase one grants or MOUs. Kesselman highlights that the model is open-source and used by programs like NATO Diana, with an AI assistant being added to answer queries and demystify processes. Regarding strategy, he notes no proven advantage to starting commercial-first versus defense-first; instead, startups should remain opportunistic, pivoting based on funding opportunities like DARPA grants.
The dual-use approach is gaining acceptance, as VCs now recognize the value of government funding and interesting problems, reducing previous stigmas. Overall, the model serves as a roadmap for startups navigating the complex defense innovation ecosystem, emphasizing communication and adaptability over rigid planning.
FAQs
The three main focuses are customer, funding, and technology development. Capturing readiness in these areas helps define a company's stage.
It is a model created to evaluate startups on three vectors—customer, funding, and technology—across two dimensions: mission and commercial. It provides a common language for startups and government to understand each other.
It helps startups understand where they are in finding and capturing customers and funding, and developing technology. It also shows the next steps, like applying for a phase one grant or getting an MOU.
AI is integrated as an assistant on the model's website, allowing users to ask questions like why they need a phase one or what a Series B round is, helping break down the language divide.
There is no empirical proof that either is better. The best approach is opportunistic—pursue customers and funding from either side as opportunities arise, and pivot as needed.
Founders can manage both by being capable and persistent. The stigma that government grants are too distracting has faded, and many startups now successfully pursue dual-use strategies.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.