How to Balance Execution and Innovation with Steve Blank

[00:00:00] Steve Blank: Innovation almost always happens at the edge. That is disruptive innovation, not mainstream innovation.

but it eventually needs to move into the center and become the core. Or else you're just entertaining yourself. The goal is again, execution pays your salary, but disruptive innovation is going to pay your pension, and so to know the short term and long term things and have different processes for them is something that we just haven't adapted and adopted yet, in the DOD at scale. 

[00:00:47] Ryan Connell: Hey, this is Ryan Connell with the Chief Digital and Artificial Intelligence Office, joined here with a very special guest, uh, Steve Blank. Steve, how you doing?

[00:00:54] Steve Blank: I'm doing great. Thanks for having me, Ryan.

[00:00:57] Ryan Connell: Yeah, absolutely. Steve, let's just turn it over to you to give a quick intro.

[00:01:01] Steve Blank: Great. let's see. Steve Blank, I'm currently an adjunct professor at Stanford University in the Gordian Knot Center for National Security Innovation. ex Air Force vet, during Vietnam. and then did eight startups in Silicon Valley in software, hardware, and, enterprise software and supercomputers, et cetera.

[00:01:20] Ryan Connell: Awesome. Uh, no, impressive, I should say. 

[00:01:23] Steve Blank: Well, I guess it makes proven to your audience, uh, helped co create the lean startup movement as well.

[00:01:28] Ryan Connell: Awesome. Okay. Let's, jump into it. I know you were super involved with, uh, hacking for defense. Do you want to just give a quick overview of what that is? 

[00:01:37] Steve Blank: Sure. Um, and you know, the short version is hacking for defense is a. Uh, class in, 60 some odd research universities in, I guess three or four countries that, take problems from the DOD and IC and hand them to students and they use a lean startup method to, to get outta the building, validate the problem, and build what are called minimum viable products.

And, uh, in 10 weeks, their sponsor and the students learn a ton. and that's been going on now for nine years since we started it. much longer version about the methodology and what it's based on and its applicability to CDAO, but it's, been a pretty successful class.

[00:02:16] Ryan Connell: Yeah, let me ask you a question because this has come up a couple times on Defense Mavericks. I'm just curious your thoughts, like, why do we always seem to have success with students? 

[00:02:25] Steve Blank: Well, you know, if, you could remember back to when you were young, you were not kind of, trapped into a set of, uh, well, this is the way we always done it, or I've done it this way for 30 years and what could be new. you know, we managed to, in almost every large organization to beat the curiosity, out of folks, only because for organizations to be large and to grow, they need to adapt and adopt processes and procedures And rules and you don't want thousands or tens of thousands of people doing their own thing. the nice part about is when you're in a university, you're when you're near a startup, you have the ability to experiment at scale. again, it's not that folks in large organizations are no longer smart, but again, the nature of the organization requires them to follow a set of processes and tracks.

so we're not all freelancing inside that organization. But without that process, without the ability to innovate, fail, experiment, learn, and discover, we tend to fall behind from, organizations that are doing that or at least built in this nature of ambidexterity, that is the ability to execute and innovate simultaneously.

I'm not sure that was your

question, 

[00:03:35] Ryan Connell: No, that was perfect. Yeah, I'm curious. Like, is there a, is there like a macro level, like if there was utopia, we would spend 70 percent of our time in the rule set in operational and 30 percent of our time experimenting and trying and failing. Is there, has there ever been like metrics put to something like that? 

[00:03:53] Steve Blank: there are, um, you know, in the enterprise, uh, in a perfect and be not perfect. But, you know, one could imagine, let me go back. There's still a model called the three horizons of innovation. I'm sure a good number of your listeners have heard it. It's an old McKinsey thing that says, you know, horizon one is kind of execution of current stuff.

And it's not that we don't innovate. We innovate along our current, products and we make them faster or better supply chain or, you know, different tool sets, etc. But still it's the same offering. Horizon 2 says we're taking the existing tools, technology, etc. and maybe repurposing them to a different customer set or, for the same customer set maybe we're using, you know, completely different solutions. Horizon three says, well, we're creating new things that never existed before. These are the crazy people. Those three horizons, um, people tend to say in a large enterprise, if you want to be ambidextrous, that is execute and innovate simultaneously, you ought to be spending.

Maybe 60 to 80 percent of your budget on execution, maybe 10 to 30 percent on horizon two, that is, you know, let's repurpose stuff and create new things, but basically same customer set or same, software, but if we're not spending at least 10 percent or more,on, budgeting for, disruptive things.

Either by acquiring them or by doing them internally, we're going to fail. we're going to fail because our competitors are going to do that. Or for the US, our adversaries are going to do that. They're going to come up with offset strategies. Or they're going to adopt AI faster. Or they're going to do X or Y.

You know, the best example of this, kind of grossly in the commercial world, is actually in hardware. though it's mostly software based, is SpaceX. NASA Jet Propulsion Laboratory, California Institute of Technology You know, on a three launch pads, SpaceX is launching every two and a half days, uh, Falcon nine, which is basically their standard rocket, which now is, you know, after 10 or 15 years is usually operational excellent, meaning, very little deviation from the norm and like there's checklists and there's process and there's procedure and you've got to work there.

They do not want you to freelance. On the other hand, in the same exact company. They have a bunch of crazy people in Texas building the next generation called Starship, and different engines, Raptor engines rather than Merlin engines. And the Elon's model is if you're not blowing things up on, in Starship, you're not innovating.

And if you're not blowing things up on a regular cadence, you're not innovating fast enough. Well, that's 180 degrees from what you want on those pads in Vandenberg and Kennedy. but in fact, it's the same company. And they understand two things. One is, the learnings are not just vertical.

That is in those silos. They're actually horizontal. The people in, uh, Falcon 9 are teaching the Starship folks. Oh, here's where we should have put the access panels, you know. Heh, we, we got that wrong. Or, here's where the ground service equipment really should plug in. We got that wrong. And the people in Starship are actually teaching the, existing system, you know, you can crank up that chamber pressure another 10 pounds and getting NASA's approval.

And so payload capacity has gone up for the Falcon 9 a couple hundred, maybe a couple thousand pounds without actually tweaking or redesigning the rocket. And I've summarized that down into a simple phrase, which is in any large organization done well, execution pays your salary, but innovation pays your pension.

And unless you have both, you end up with a. n organization that, you know, is innovative, creative, grows, scales, and then gets disrupted by a new organization, you know, a competitor, or again, in our case, an adversary, who's on the next wave of technology. And so, this is just the nature of both organizations and the adaptation or not of new technology waves.

You know, the current one might be AI, but you know, might be 15 other, pieces of tech or its ability to kind of simply adapt and adopt at the rate of change.I'll take a full stop there, but you know, the, the nice part is CDAO has a, pacing set of systems that they should be measuring themselves against.

One is obviously the rate of change of, software and systems in China, but the other is Silicon Valley, you know, if we're not moving as fast as either of those two, you know, the consequences are fairly large.

[00:08:17] Ryan Connell: Appreciate that. Um, so something you said, I jot it down here. I I'm curious, you know, I'll call it the state of innovation or maybe in your words, how many crazy people do we have? But, in the department, like, you've been at this for years at this point. I'm curious if you've seen, the increase in crazy people, like, are we, breaking stuff? Blowing stuff up fast enough? or maybe fast enough isn't, the right question because you just talked about how to pace it. maybe I'll ask it differently. Like, has it improved?

[00:08:44] Steve Blank: Well, you know, in an organization as large as the DOD, there are certainly pockets of hope and, actually, incredibly innovative people. and there are also pockets of people who think nothing's changed in the last 30 years and everything's just fine. you know, the one I'm kind of, at least currently, most intuitively familiar with is,the Navy, the active CTO, Justin Finnelli, and he's the PO of digital. it's probably as good as anything in the, start up world. And that doesn't mean it's as, the Navy's adopted things as good as anything in the start up world. But at least that's a group that, from where I sit, and I just might be biased, gets it and is adopting and adapting as fast as the palm cycle will let you do that.

so yeah, there's certainly large chunks of hope. There's also large chunks of, you know, unhope where people, and this is just the nature of a large organization. We have the largest organization in the world. Where people think their title is their job. And what I mean by that is that, Oh, what am I responsible for?

Oh, I'm responsible for acquisition of X, or requirement Y, or what? No, not really. You think you are, but that's actually a denial of service attack on the country. Your job is to facilitate, if you're an acquisition, You know, a new capability that makes the country safer and more secure and do it within the bounds of what's legal and, you know, what the acquisition rules are.

But it's not your job is that piece of the process. Your job is to facilitate that flow of the process as rapidly as possible. and the problems I typically run into or see are. People who believe their job is not passing that paper from one end of the desk to the other. Not understanding that, no, that's actually an obstacle.

Rather, they've become an obstacle rather than a facilitator. Does that make sense?

[00:10:36] Ryan Connell: Yeah, a hundred percent. And it reminds me of one of our cultural tenants here is, uh, if not us, then who? In, in this mindset of like, you have to understand that whole process and, all of the processes, uh, to facilitate that. 

[00:10:47] Steve Blank: Right.

And, and the other part, you know, and I was just, thinking about this and, and wrote a blog post about this, about why is adoption of new ideas really so hard? and new technologies. And, you know, if you're an innovator, you tend to look at large organizations with the biggest obstacle being the frozen, what's called the frozen middle, you know, people don't want to change and whatever, and, you know, spending time on it.

And I've actually broken that down to, you know, it's a little more complicated than that. Number one is that there really is a part of normal human nature. That's called the Semmelweis effect that was kind of observed a couple hundred years ago is that for a good chunk of people, I could put evidence right in front of you.

But if your head is wired to believe X, you will go, yeah, that's nice. And go back to believing, you know, what you currently believe. And we can talk about how to fix that, but that's a good chunk of the frozen middle is the Semmelweis effect. It's not out of malice. It's just an unconscious bias to ignore evidence.

It's a big idea. We could train people to not do that, but unless we understand, that's just normal wiring. Because if we were flip flopping every day when we saw something new, we'd probably get eaten by lions and tigers on the savannah. The second part is, and the second obstacle is a conscious thing, which is kind of a little bit of malice, which is hubris.

Which is, gee, I'm in charge of X, and my ego is wrapped around the number of people, or the budget I run, or whatever, and You know, if you're challenging that, or you got a better idea, but it might take away budget from my program or something else over my dead body. Well, you know, that's a leadership failure, of being able to manage that.

So, so we got unconscious bias. Some of us, in fact, we've got conscious bias. Then the third one is, which is only a government problem, but it is a problem. And I'm sure we've all seen it, which is the short tenure of leadership. Right? Someone appointed to a program, you know, likely has a two to three year tour, certainly if they're an active duty officer.

and then the third, and by the way, that's just the nature of the system. I don't know if we're going to change it, but it does not occur in the civilian world. Right? In a civilian world, you have, you know, as long as you're succeeding, you stay in that job. It's worse. If you're succeeding, you might get promoted the job.

I actually learned something. We're making a difference. And usually for large organizations, if you want to do, um, substantive certainly transformation two to three years, just, you're just warming up. The other,government only kind of, problem is, the revolving door with, with industry.

All right, there's no way you're going to blow the whistle on a major contractor If in fact you're within, you know a decade or two of retirement because you're thinking that there's no way I'm ever going to work in that industry if I call someone out for shoddy work or not delivering or whatever.

you know, that shouldn't happen and we should have better checks and balances and whatever. but it happens a lot and I'm sure, you know, people inside the, agencies and organizations have seen that happen. that just kind of has a drag on, the rest of the systems because we don't hold our vendors to the same accounts you typically do in the civilian world.

And then, of course, one of the biggest obstacles is our own acquisition system, you know, tied to, acquisition with a big A, the JCDIS, and the budgeting process, and the writing requirements, etc. getting stuff into the POM, and, you know, that's too, you know, think about it. You know, a startup could become a unicorn in the same time it takes to get somethinginto the pub.

I mean, that's insane, right? I mean, though we now have shortcuts of OTAs and being able to acquire commercial stuff, et cetera. But still, if we're a traditional acquisition person and we want to go do something large, you know, the time lag, I mean, you know, if we think about it, we're still putting stuff in.

Into Congress before chat GPT, you know, like, Whoa, Hey, there's this new thing called transformers in general of AI. Oh, well, that's not what we're requiring. Cause that wasn't around, you know, three years ago. I mean, literally think about it, the whole AI thing at scale, it's three years old,

[00:15:04] Ryan Connell: I mean, I, know a very, very intimately requirements that were like in place in January and OBE by April. yeah, a hundred percent. 

[00:15:11] Steve Blank: And then the third of is the, both the reluctance and institutional, I don't know what it is, not to get out of the building and spend a lot of time, kicking the tires of, new entrants, particularly, and when I say Silicon Valley, I don't mean the physical place, I mean from innovation clusters and new entrants.

I mean, you know, if it were me, I'd like have a, twice a year quiz for anybody in acquisition of tell me who the newest players are and, how do they match the requirement to writing, because there's no way you could do this if you haven't gotten updated every six months, especially in software, it's insane that you really need to be constantly engaged.

And that not only isn't in our culture, you know, that isn't mandated. I mean, they'll There are laws that says we're supposed to get out and look for best available, etc. but we don't have those processes to do that.and then, finally, the lack of a fast track for, disruptive things, innovative things at scale that are non standard.

One, if you remember Horizon 1 and 2, we have lots of paths for that and for those people. But in most parts of the DoD, We don't have a fast track and innovation pipeline for disruptors, right? In fact, disruptors get, slapped around, knocked down and not promoted rather than say, Oh, crazy people.

You go over here, and we have essentially, you know, appendix a rules for you guys, which are, you know, you turn to the back of that the rules manual and there it is for the small group and this goes back to are they special? Well, no, they're special like starship is special that there are different groups working at different clock speeds with different outcomes with different cultures. You cannot put a disruptive Group inside of the mainstream of your organization, or it will get crushed for, Oh, we need the budget.

Oh, we need the people or G it's the wrong people, wrong culture. These need to be separate organizations with their own rules that eventually tie back to the mainstream, because what happens is innovation almost always happens at the edge. That is disruptive innovation, not mainstream innovation.

but it eventually needs to move into the center and become the core. Or else you're just entertaining yourself. The goal is again, execution pays your salary, but disruptive innovation is going to pay your pension, and so to know the short term and long term things and have different processes for them is something that we just haven't adapted and adopted yet, in the DOD at scale.

[00:17:40] Ryan Connell: Got it. I'm going to, I'm going to poke a bit. Uh, I think the first of several things that you just brought up, I'm curious to your thoughts. I don't know that I've ever said this out loud, but it's been running through my head, um, for years at this point, but with the, frozen middle, concept.

Uh, I think that was your number one and number two is, you generalize it's a leadership failure.I've thought a lot about the idea of the frozen middle and how you have kind of those senior execs that are, yeah, yeah, let's do it. Like jump on board, really engaging, uh, inspiring to the junior workforce that wants to innovate, make new ideas.

Then you have this middle that's, not willing to move and I've felt like, Well, isn't it the senior executive that that role's job to push authority down, put, you know, like, push information up like, it's one of those things that I view that as potentially also a leadership failure. And I'm curious what your thoughts are on that.

[00:18:30] Steve Blank: Well, you know, one of the problems are, and this happens to everybody over the age of 21, is you tend to remember the things you were taught, you know, in your twenties.your first, you know, your first boss, or your first job, or your first, training sessions, or whatever. As you move up the ranks, the world changes underneath you.

but the amount of time you have for learning, not your ability, but the amount of time you have to learn new things and kind of experiment, et cetera, you're now into managing. And so you kind of get stuck with the frameworks and tool sets you had 10, 20, sometimes 30 or 40 years ago. it's not just the technology has changed, but the methodologies of how to innovate have changed.

And the problem with most senior leaders inside the DoD, is they simply don't know what those are. It's a big idea. If you ask most of them is, well, how does Silicon Valley innovate? What's, what's the lean startup look like? And, and how do they move? How does software get develop so fast? And, no, no, no, don't tell me it's all about the DOD.

Tell me about how does the Valley do that? And again, Valley, I mean the, not just the physical place, but innovation clusters. The problem is our senior leadership needs to be educated, not first, they need to be doing these things first. We need to kind of upskill them about, well, what are these methodologies?

and I'm not just talking about, DevSecOpsand, you know, software development tools. I just mean innovation methodologies at the highest level and because we need to remember China is operating like Silicon Valley in a good day. You know, the D O D operates like General Motors.

That's just an unfair. It's not that we don't have a smart people. It's that we're operating with, mental models that are decades old. Does that make sense about leadership? 

[00:20:11] Ryan Connell: Yeah, absolutely. 

[00:20:12] Steve Blank: So. So, yes, to your point. Yes, this is what leadership should be doing, but they simply don't know. Not only don't know about the specific tools that have changed, but about the specific methodologies of software development have changed.

And then this different all the way on the top, different methodologies about thinking about rapid innovation have changed. And you could do that within the framework of the POM and the, and PEOs and whatever, and figure that out. But that's. And what's really interesting is, of course, is that this is the other problem in large organizations and DOD writ large, is that you have innovators on the bottom and the middle who know all this stuff, but they're not connected to senior leadership.

What's in between them and senior leadership? The frozen middle. So, you know, a cheap hack would be to have, you know, the most senior leaders get briefed once a month by the crazy people and some of the adults who are managing the crazy people, to say, Sir, here's what's going on. 30 seconds. Boom, boom.

Boy, that would just, I mean, you don't even need to go to school for that. You'd be going, wait, we got where, we got what? Where in the bottom of our organization? Well, why isn't it? Oh, well, it's not budgeted. It's not scheduled. That's not whatever. I mean, that's the problem when you have a single track is you could invent anti gravity and some mid level manager will say, well, that's not on my budget and schedule this year.

Go back to work on what I've assigned to you. Trust me, I've actually seen that. Multiple times. And again, there's no malice involved. It's that the process dominates, innovation and unless you have a active culture to, to bring out those rapid innovation changes and figure out how to integrate them, I'll give you a good example.

I won't, Mention the agency, but because it was a failure on my part and leadership's part is we stood up something called I Corps inside of a major part of the DOD. And we had innovators essentially doing TDY inside their organization for six weeks working on, you know, like real problem sets, but outside their org.

At rapid speed building minimum viable products. That is things that could actually be used if they were budgeted and scaled. And then they went back to their organizations. Okay. Here. and again, their middle managers went, well, what the hell is this? It's not on my budget. it's not scheduled.

Get back to work. And we literally trained thousands of people on this rapid innovation process, but we never had leadership adopt this. and without that, without some alternate pipeline and some, you know, 10 percent overhead on budget that we could figure out how to get. Basically the capabilities organization said, Well all this innovation is a denial of service attack on our core capability.

And they were right because we hadn't figured out how this fit into the entire flow. It was literally throwing stuff over the transom. So just getting the innovators engaged without getting the organization designed to, Be able to take some of this innovation and focus it in a funnel and a pipeline that's parallel to our existing work.

That is, we, again, this ambidextrous organization nature, really comes to fore again.

[00:23:31] Ryan Connell: Yeah. And I think I think that cheap hack that you mentioned isn't as cheap as you think. I think that there is, um, some true value to that, even if it's just to your point earlier, like the awareness exec, uh, that general or the,state of uh, Uh, state of art is or state of the possible or state of technology.

Right? Like I'm just thinking through all of the conversations that the junior workforce probably isn't in terms of, we got, you know, a hundred K of expiring funds. Anyone have any use for it? And everyone's like, no, no one And if, only they'd And if, if only they'd had that conversation and realized they could have done something with 10 or whatever the they, had that because they, they had that 

[00:24:08] Steve Blank: And, and, you know, let me give you a specific, if I wanted to set this up inside, I'd set it up the same way I would do a tour outside the building. And one is instead of just showing me your shiny object, I wanna remind you, here are the top problems that senior leadership says we're, we have great, I don't wanna see a shiny object of giving me a, a AI demo.

Give me an AI demo in some form of operational context that matches the problem sets I've just told you I'm worrying about on the seventh floor or the floor with a thick carpet, right? . Show me those, and that's what I wanna see once a month. Now all of a sudden I'm connecting, you know, the lowest levels of the org, or at least the mid levels of the org, as a reminder, here's what I'm worried about up here.

Got any solutions? And, and it's the same when these senior folks visit Silicon Valley or technology clusters. They tend to do, you know, an innovation theater tour. Oh, here's shiny building, or here's demo X or Y. Well, that's a waste of everybody's time because what they should have done is have their staff socialize the problem sets with those companies.

They're going to go visit and say, I don't want to see a demo of your standard package. If you're interested in working with the government. I want to see a demo as applied to these sets of problems, or how would you solve these problems? And then I will bring, you know, like our SES or general or admiral through this tour of companies, and now you could do a debrief of going, Oh my gosh, our head's exploding.

Look, they have an answer for X or Y. We could buy this off the shelf. And of course, your contractor is going to melt down when they see that because they, you know, but, that's the nature of, how we should connect those ends, not just shiny object from both inside and outside, but problem focused, particularly if you could do it in some operational context, does that,

[00:25:59] Ryan Connell: Yeah, and that resonates so well for me. I hope it does for the listeners too. I mean, I'm just thinking through, all of the organizations that publicly display their lines of effort, strategic engagement, lines of effort, all of the things that they're trying to tackle and, you know, for the vendor pitches for, the, pitch ins for, you know, subordinate to supervisor or for industry to, government, to your point, explaining how, what they're offering.

Let's that person solve their own problem. that's so impactful. Yeah. all right, I want to jump into, um, you've mentioned it a couple of times now. What is the lean startup methodology?

[00:26:33] Steve Blank: well, it's, you know, by now in Silicon Valley, it's kind of like second nature, at least it's second nature to kind of, Mention it. it's actually sometimes hard to do, even though it's there's no technology involved. It's a Lean Startup came from my observation, that said, and we will translate this into writing requirements.

When we did startups, we would basically, you know, the founders would go, I think we ought to build X. And the, I think we ought to build X was based on either their experience in, a problem they wanted to solve in their last company or, their notion that a new technology could be, you know, Perhaps useful in a, commercial setting.

But basically we get together in a conference room where, you know, write a spec for engineering and we'd try to raise money and then we'd go build the prop and there was almost zero customer input. I mean, we'd maybe go through the waterfall development process of, you know, functional spec, alpha test, beta test, first customer ship.

But the only reason we'd be going through alpha and beta is to test for bugs and stability of the product. Not that we cared whether customers wanted it or needed it. It was like, oh, that was what we were building. and then surprise, we would ship it expecting that we hopefully had an office big enough to hold all the bags of money.

and it never worked out like that. It was either too late or You know, people didn't want all those features or we missed the key features. So we were missing what was called product market fit with that is the fit between the customer segments and the value proposition, meaning the features or service or, whatever, we didn't quite understand that because number one, we assumed we understood the problem.

And because unconsciously we assumed we understood the problem, we consciously decided to build a solution without checking with anybody else. Does this ring familiar in writing a requirement? 

Right? Like we still make in the DoD, and then we would build it in a serial process, step by step, using waterfall.

So the lean method said, that's insane. Why don't we use waterfall? Three things, just three things. One is, why don't we assume that all we have on day one is a hypothesis? Hypothesis about the problem and solution. So why don't we get out of the building and actually talk to potential customers, in this case warfighters, stakeholders, regulators, etc, and validate that what we think, that just what we already thought.

We think we're right, let's go validate, and also the time that they needed. Number two is, and this was, so that was my invention, was called customer development. But at the same time, what emerged in the 20th century, or in the turn of this century, was Agile development as an alternative to Waterfall.

Agile simply says, no, don't build things in a serial process, because you only discover you're wrong at the end after you spend a lot of time and money. Even though that's how we write requirements and do acquisition, for, just hold that in abeyance for a second. What if it was possible to build the product?

iteratively and incrementally is while we're getting out of the building. Why don't we show those customers something called a minimum viable product? It could either be a wireframe or a spreadsheet or a screenshot or something that allows us to get feedback, not only about their needs, but about whether we're built in the right thing.

Well, in the commercial world, this allows us to do something called a pivot. A pivot says, gee, when you get evidence that what you're building is wrong or wrong customer, wrong feature set, wrong timing, you have permission to change. You could change like product features or you could change who you're trying to sell it to.

You could change timing or you could change cost. So that you have permission to do that. And that pivot was incredibly powerful because it says instead of delivering things that no one wants or needs or is wrong or whatever. We actually could better hit the target much closer to a hundred percent. And if we didn't, it's because we didn't listen.

And then the third is we just use something called the business model canvas or in the military mission model canvas to kind of track all the things we need to be thinking about, you know, do we have the right customers or have we talked to all the, in this case, the users are almost never the buyers.

And the same is obviously true in the D O. D. Right. Their users and buyers. And now we got Congress and we got staffers and we got requirement writers and we got program. But so, so when I'm talking about product market fit, it needs to fit all those things. And all those folks have different needs, right?

The users want X, but the buyers want Y. Oh, and by the way, in the D. O. D. We also have saboteurs. Saboteurs are, you know, the existing owners of the contract, right? Or the existing whatever. If you have something new, it might not be from them. so the lean method allows us to develop things rapidly with speed and urgency and deliver them quickly and hit the target of not 100%, but certainly a lot better.

Than trying to pre compute the future, and by the way Waterfall worked in the 20th century because we were competing for the soviet union then we're essentially using the same methodology But now we're kind of out of sync as I said seriously China is using, you know, lean, this lean method is how China's putting, you know, how many DDGs or disorders that they've put in the ocean in the last, you know, 10 or 15 years.

How quickly are they modifying their systems? Why are we stuck? It's not, we just can't keep blaming the palm. We also should be thinking about why are we still like thinking, if we're still thinking about waterfall, why on earth are we doing that? We now have some contracting methods that allow us to start thinking about.

How do we deliver things with actual contracts that allow us to change and set of requirements for every feature to actually deliver outcome based contracts? Or deliver service based contracts or deliver, you know, why don't we just buy the solution or the outcome rather than buy each piece of the software as a checklist.

and the thing I keep asking requirements writers is, do you really think you could predict the future for every possible feature? Because no one has been able to do that. Let me know if you're going to be the first one. And that's what I could go back to. Oh, well, that's my job. No, it's not. Your job is actually to deliver a capability to a warfighter or, you know, someone who needs a system, not to follow each checklist blindly like this is the only way to solve this problem.

Does that make sense?

[00:33:04] Ryan Connell: 100%. Yeah. I mean, if you're an innovative solution provider, it's almost like a, how dare you put me in a box? Right. Um, Hey, curious. I got to ask. So, so we talked a lot about innovation innovators. Um, do you have a, Steve blank definition of innovation?

[00:33:22] Steve Blank: Well, you know, to me, innovation is the creation of a,capability or product or service that never existed before.and again, it could be as simply as, gee, I added a new feature to something that we're already building or, figured out how to get it cheaper or whatever, and again, you know, the, language is that sustaining innovation.

it's not any less valuable than disruptive, but it's creation of something new, disruptive innovation is again, the creation in this case of something that never existed before, but I also wasn't building before. It wasn't a product or services we had, but all of a sudden we realized it might be incredibly valuable.

and again, let me give you a great example is, in software, it was Amazon. Amazon was doing innovation all the time. If you looked at their core business of, you know, their website of selling books and then other things, they were constantly innovating about, Oh, well, we started with books and now we could sell you diapers and we could sell you something else.

And that was great sustaining innovation. And then the horizon went into. But AWS, AWS was not in their business at all, right? Like, what? 

You know, well, right, AWS or the equivalent nowadays, whether it's Azure or someone else's, is basically, you know, electricity. It's like a utility. Like, but that came from a company selling books online.

That's an example of disruptive innovation, right? That was a creation of something that never existed before outside your current business model.

[00:34:51] Ryan Connell: Awesome, I appreciate that. Hey, getting close to wanting to wrap here. I'll just turn it over to you if you have any, um, anything I didn't ask or you don't want anything that you wanted to share to close out

[00:35:01] Steve Blank: Well, your listeners should just know that at Stanford, we started something called the Gordian Knot Center for National Security Innovation, and that was our observation about four years ago that And we're kicking ourselves that we should have done it much earlier. That here we are, and this time I do mean physically, at Silicon Valley.

Proximate, that is next to the creation of people doing AI and, you know, autonomy, and like semiconductors and enterprise software. I mean, 90 percent of it sits within 50 miles of Stanford. Yet we also had world class policy people, world class engineering people, world class faculty in policy and engineering.

And there's a good chunk of the campus that deeply understands the DOD. But there was no center that integrated all those pieces. And so the Gordian Now Center, basically said, Look, why don't we, start a center with a series of classes and innovation fellows that focus on that area? Because it's pretty clear, and I think all your listeners understand, that for the first time ever in the history of the United States, the DoD no longer owns, all the technology necessary to win, or deter a war.

I mean, that was just never the case. It used to be our FFRDCs, our federally funded research and development centers, or our PRIMES, or our service labs, all had every piece of technology we needed, which was more advanced than anyone else. I think everybody understands that's no longer true. I mean, obviously nuclear, hypersonics and Exquisite sensors still belong in the DoD But if you're talking about cyber and AI And, autonomy and go through the checklist, you buy that off the shelf. Go ask the Ukrainians right 

about drones. 

[00:36:41] Ryan Connell: Yeah, absolutely.

[00:36:42] Steve Blank: Um, so this notion of actually having a center that's worrying about those things and thinking about their implications and teaching both engineering students and policy makers, how to deal with that is why we created the center.

So there's about 40,research for national security innovation fellows. And then we teach both hacking for defense and a class called technology innovation and great power competition. and Joe Felters, the executive director, Eric Vollmer, who was, number two at the Office of Strategic Capital, Raj Shah, who was, number two at, or the second head of Defense Innovation Unit, was a co founder, is now at Shield Capital.

And I teach there as well. So that's the, that's what the center does.

[00:37:21] Ryan Connell: Awesome. Yeah. I appreciate it. Well, Steve, appreciate you, being on today. let's go earn our pension. Uh, I'll do the same, in terms of our innovation earned pension. Um, but yeah, I appreciate you being on a excellent conversation.

[00:37:33] Steve Blank: Great. Thanks for having me. It's always fun, Ryan.