In the Realm of the Barely Feasible with Arati Prabhakar [Idea Machines #37]

Idea Machines

In this conversation I talk to the Amazing Arati Prabhakar about using Solutions R&D to tackle big societal problems, gaps in the innovation ecosystem, DARPA, and more.

Arati’s career has covered almost every corner of the innovation ecosystem - she’s done basically every role at - DARPA she was a program manager, started their Microelectronics Technology Office, and several years later returned to server as its Director. She was also the director of the National Institute of Standards and Technology and was a venture capitalist at US venture partners. Now she’s launching Actuate - a non-profit leveraging the ARPA model to go after some of the biggest problems in American society.

Links

Actuate Website

In the Realm of the Barely Feasible - Arati's Article about Actuate and Solutions R&D

Arati on Wikipedia

Transcript

[00:00:00] welcome to idea machines. I'm your host and Reinhart. And this podcast is a deep dive into the systems and people that bring innovations from glimmers in someone's eye, all the way to tools, processes, and ideas that can shift paradigms. We see these systems outputs everywhere, but what's inside the black boxes with guests.

I dig below the surface into crucial, but often unspoken questions. To explore themes of how we enable innovations today and how we could do it better tomorrow.

In this conversation, I talked to the amazing RFE provoca about using solutions R and D tackle, big societal problems, gaps in the innovation ecosystem, DARPA and more. Are these career has covered almost every corner of the innovation ecosystem. She's done almost every job at DARPA where she was a program manager, started their micro electronics technology office. And several years later returned serve as their [00:01:00] director.

She was also the director at the national Institute of standards and technology and a venture capitalist at us venture partners. Now she's launching actuate a nonprofit leveraging the ARPA model to go after some of the biggest problems in American society. Hope you enjoy my conversation with Arthur. Provoca.

I'd love to start off and sort of frame this for everybody is with a quote from your article, which, which everybody should read and which I will link to in the show notes. You say yet, we lack a systemic understanding of how to nurture the sort of rich ecosystem we need to confront the societal changes facing us.

Now over 75 years, the federal government has dramatically increased supportive research and universities and national labs have built layers of incentives and deep culture for the research role. Companies have honed their ability to develop products in markets, shifting away from doing their own fundamental research in established industries, American venture capital and entrepreneurship have supercharged the startup pathway for commercialization in some [00:02:00] sectors, but we haven't yet put enough energy into understanding the bigger space where policy finance and the market meet to scale component ideas into the kind of deep and wide innovations that can solve big previously intractable problems in society.

These sorts of problems, aren't aligned to tangible market opportunities or to the missions of established government R and D organizations today, the philanthropic sector can play a pivotal role by taking the early risk of trying new methods for R and D and developing initial examples that governments and markets can adopt and ramp up the hypothesis behind actuate is that solutions R and D can be a starting place for catalyzing the necessary change in the nation's innovation ecosystem.

And so with that, with those, I think I want to test it in a nutshell exactly like that. So can we start with how do you see solutions R and D as being different from other R D and, and sort of coupled with that? How is actuate different from other non-profits. Yeah, I think [00:03:00] that's, that's one of the important threads in this tapestry that we want to develop.

So solutions R and D let's see. I think those of us who live in the world of R and D and innovation are very familiar with basic research. That that is about new knowledge, new exploration, but it's designed all the incentives, all the funding and the structures are designed to have that end with publishing papers.

And then on the other hand, there's. But the whole machinery that turns an advance into, you know, takes a technological advance or a research advance and turns it into the changes that we want in society that could be new products and services. It could be new policies, it could be new practices and that implementation machinery.

The market companies, policymaking, what individuals choose to do pilot practices. I think we understand that. And there are places where the, you know, things just move from basic research over into actual [00:04:00] implementation. But in fact, there are, there are a lot of places where that doesn't happen, seamlessly and solutions, R and D is this weird thing in the middle.

That builds on top of a rich foundation of basic research. It has it, its objective is to demonstrate and to prove out completely radically better ways. To solve problems or to pursue different opportunities so that they can be implemented at scale. And so it has this hybrid character that it is at the one on one hand, it's very directed to specific goals.

And in that sense, it looks more like. Product development and marching forward and, you know, boom, boom, boom, make things happen, execute drive to drive, drive to an integrated goal. And on the other hand it requires a lot of creativity, experimentation risk-taking. And so it has some of those elements from the research side.

So it's this middle [00:05:00] kingdom that I. Love because it has, I think it just has enormous leverage. And I, you know, I, I think a couple of points, number one, it's it requires to do it well, requires its own. Types of expertise and practices and culture that are different from either the research or implementation.

And secondly, I would say that it, I think it's overall in the U S in the current us innovation system. I think it's something of a gap. There, there, there, there, there are many, many areas where we're not doing it as well as we need to. And then for some of the new problems, which I hope we'll talk about as well.

I think it's actually a very interesting lever to boot the whole system up that we're going to need going forward. Yeah. And so actually just piggybacking right off of that, you've outlined three major sort of problems that you're tackling initially. Climate change sort of health, like general American [00:06:00] health and data privacy.

I'm actually really interested in, like, what was the process of deciding, like, these are the things that we're going to work on. Yeah, but this whole actuate emerged from a thought process from a lot of. Bebe's rattling around in the box car in my head in the period as I was wrapping up at DARPA in 2016, at the end of 2016 and going into 2017 when I left and what I was thinking about was how phenomenally good our innovation machinery is.

For the problems that we set out to tackle at the end of the second world war, that agenda was national security technology for economic growth. A lot of that was information technology. We set out to tackle health. Instead we did biomedicine. We went long on biomedicine, didn't break their left, left a lot of our serious health problems sitting on the shelf and a big agenda was funding, basic research and, and we've executed on that agenda.

That's what we are [00:07:00] very, very, very good at what I couldn't stop thinking about. As I was wrapping up at DARPA is the problems that I think will, you know, many of us feel will determine whether we succeed or fail as a society going forward. So it's not that these challenges, you know, national security or how it's not that those problems have gone away and we should stop.

It's just that we have some things that will break us at our. Yeah, arguably, they are in the process of breaking us. If we don't deal with them right now, one is access to opportunity for every person in our society. A second is population health at a cost that doesn't break the economy. Another is being able to trust data and information and the information age in which we now live.

And the forest obviously is mitigating climate change. And if you think about it, these, these were not, but these weren't the top of mind issues at the end of the second world war, right? I mean, we had other problems. We didn't really know what to do about. So some of these are all problems that we didn't really know what to do about.

Some of these are new problems. And, [00:08:00] and so, you know, now here we are in 2021, if you say what's what really matters those were the four areas that we identified that. Are critical to the success of our society. Number one, number two, we aren't succeeding. And that means we need innovation of all different types.

And number three, we, we don't, we're not innovating, you know, we're either innovating at the zero billion dollars a year level, or we are spending money on R and D, but it's not yet turning the tide of the problem and, and that, so that's how we ended up focusing on those areas. Got it. And what could you actually, like, I, I love digging into sort of the nitty gritties of like, what was the process of designing these, these programs?

Right. So just to sort of scope this a little bit, these broad areas that I'm talking about, I think of as. But the major societal challenges that we face today, actuate, which is a tiny early stage seed stage [00:09:00] nonprofit organization. Our our aspiration is over time to build portfolios of solutions, R and D programs.

In each of these areas. And so very, you know, you, you, you made reference to a couple of the specific programs. One is about being able to access many more data sets to mine, their insights by cross-linking across while rigorously preserving privacy. That's some of the whole set that's one very specific program, but, but think of that as just one program and what will eventually be a much broader portfolio in this area of trusting data and information.

So part of what we've been doing as we started actuate in late 2019 was big thinking about our strategy, about the four broad societal challenges that we wanted to work in. And then we've also been doing a lot of work on we've defined a couple of specific programs, but perhaps more importantly for scaling the organization, we've been working through our [00:10:00] art.

Our mess, our process and methodology to take, you know, the core idea here of course is our, our founding team has a lot of different experiences, but we met at DARPA and what we our inspiration is really to take what we know from that particular model for solutions R and D. And. Mine, the critical, the essential insights and translate them to these very different societal challenges, not national security, but the ones that actuate is gonna focus on.

And, and that, so we've, we've been formulating the four areas, but also thinking through, so how do you get from the question of changing population health outcomes to what are the programs that could be high leverage opportunities to do solutions R and D for that objective? Yeah. And so, so there's, there's sort of like two steps.

There's one is like going from like the broad area to a specific program. And then there's another, which is sort of designing the [00:11:00] program itself. And I'm interested in what, what w what do you actually do to design the program? Like what, what is, what does that look like? Yeah. Go ahead.

The first two programs that we have built out and defined were developed, were invented and designed by my co-founder Wade Shen he was a DARPA program manager for about five years. That's where we met his areas artificial intelligence and data science. And if you work in that area, you can work on any of the world's problems.

And he, he worked on an amazing array of different problem areas as well as. Programs that at darker that drove the AI and data science technology itself forward. So you know, DARPA is a building full at any moment in time full of it. It's got a hundred amazing program managers in it. Wade was one of the really exceptional one people even in that very elite crowd.

And so you know, Wade can And this is how he [00:12:00] thinks about the world. As you know, we came together because we share these concerns about these major societal challenges and a passion for bringing this kind of solutions R and D to these problems. And then Wade is the kind of guy who can invent these programs, you know, like he can just go do it.

He knows how to think about it. He knows how to go do the research and talk to people and line up a program that could really be very impactful. So we, we weighed spelt these two programs, partly because we wanted to understand what that looked like in these areas. And but you know, that's the, as we go forward, we're going to need a process that engages a community of different people.

Because over time, we're going to want to build our cadre of program leaders who will define, and then execute the solutions R and D program. And by definition, they can't all be, you know, they all, they can't all be weighed, right. We need to be able to draw from the talents and insights and the passions.

With of people who have all kinds of backgrounds technology backgrounds, deep research backgrounds lived experiences [00:13:00] on these problems. People who have, who really, you know, deeply understand how the systems work that create opportunity or population health or, or take away from those objectives.

And so a lot of what we've been doing is figuring out. So that's the question I was, if you want to change the future of health in the U S so that instead of spending twice as much as other developed nations per capita on healthcare, and yet having dozens of other countries that have longer lifespans and lower infant mortality rates, which is just criminal for the world's richest economy, if we want a future where that is radically different, where we don't have a hundred million people who either have diabetes or at risk of diabetes, where we don't have.

Can, you know, we don't have a public health system. That's thoroughly incapable of containing and disease. Like COVID-19 unlike many other countries around the world. If we want a different future than you know, That's the landscape. And how do you get from that broad set of what we want to, [00:14:00] what do you do about it?

And I think what that process looks like, so it has a top-down part and then a bottoms-up part. So the top-down part is understanding that landscape it's, it's the kind of, you know, it's understanding what, how big the problem is. What is the nature of the problem what's who's doing what I mean, these are big complex systems, right?

There are many, many, many different kinds of actors. Actors practices, culture that you have to understand. You have to have some notion of how all of those complex systems components are operating and interacting. And then you can start thinking about where there are gaps or opportunities, but still at a very strategic, broad level.

And that's about it for top-down because then of course, the model emulating a lot of the power we found in the way DARPA works is then to flip it, to bottoms up. And so then we go find people who are experts. In some aspect of this, again, they might have deep research expertise, deep knowledge of the specific problems or the way the system works.

What you want is people who either [00:15:00] know or are willing to go learn enough about what the boxes, and then be willing to live outside of it and figure out how to recast it in a different way. And And, and then, you know, similar to DARPA, there's a process of nurturing and coaching, but allowing these smart individuals to bubble and brew program concepts from, you know, like a couple of bullets on a chart eventually to a full executable program, you know, a process that I think even for someone who's super good at this take six months or a year.

So that's what we're just starting to embark on. Got it. And so that's sort of the beginning of of programs. I'm also interested in sort of like, What you hope to happen at the end of them? Sort of you're, you're in a slightly different position that DARPA, which sort of has a, hopefully a way being customer in the DOD.

That's one of the funniest ideas on the planet. I just love it when people say, Oh, [00:16:00] well, It's easy because DARPA has DOD waiting for it.

All right, please. Yeah. Let's, let's talk about how, yeah, I, okay. So yeah, let's, let's talk about that. And, and yes. And then what do we do? Right. So at DARPA, I first of all, think about six decades of history at DARPA in. Two halves for across generations of that agency. About half of what it has done is prototype military systems, things that were just crazy, that the services would never have tried by themselves, but were very directed at a specific military platform or capability.

The other half has been. Sparking core enabling technologies. And that was out of a recognition that if you build your new military capabilities out of just the same old ingredients, you're only going to get so far and you need some very disruptive core technologies. So what came out of military systems?

Iconically, of course, it's stealth aircraft. There's a much, much, much longer list, but that's the [00:17:00] easy one that everyone knows. A lot of people know that story in the national security world. Of course, what came out of core, enabling technologies. Well, arguably the entire field of advanced material science, but also ARPANET and the internet the seeds of artificial intelligence, advanced microelectronics, Microsystems, huge numbers of technological revolutions.

So if that's what's going on at DARPA the first thing to point out is that half of it and some of the most transformative. Core technology, things that have come out of DARPA did not transition to the world because DOD went and bought a bunch of it. Right. And so and, and so the transition for most of the core enabling technologies is out to industry to turn into products and services.

And, you know, we've seen. We've seen many, many stories and how that works often, what it looks like is a project that darkened funds at a university and or company. And then those individuals beyond DARPA funding go forward, identify markets, raise capital, build businesses, [00:18:00] build product lines, build industries, changed the world.

Right? So we, that that's that that's not trivial. In itself. And then, but I think I just want to also be clear that even for the half of DARPA, that's been about building prototype military systems by and large DOD is not excited about someone they'd start. I'll tell you just, just one story. When I came to DARPA, we had just started just before I arrived, we had started a program.

A great partner manager had been a Navy officer. I was serving at DARPA and he said, you know, wouldn't it be great if the Navy had an autonomous vessel, a ship that could leave the pier and navigate across open oceans for months at a time without a single sailor onboard, not a remote control vehicle, but one that just had sparse supervisory control, radically different tools for the Navy, if something like that existed.

And maybe we can actually do that. And the Navy got when the DARPA was. Trying to do this and the Navy thought, but I observed this. And what they thought was that is a [00:19:00] really bad idea. And they tried to shut it down there. Important element of DARPA is the Navy doesn't actually get to tell their people what to do.

And my predecessor appropriately said, I don't know if she said, thank you. But she definitely said, we're just doing this. By the time I got to DARPA, the Navy had gone from outright hostile to merely deeply skeptical, which is pretty important because that's the stage. It was. People will tell you what, you know, all the reasons that they don't believe it.

And they say, well, how is it going to meet call rags, which are the rules of the road for navigating, you know, in dense areas. And how's it gonna last that long at sea in that harsh Marine environment, they had the entire long, difficult list of challenges. So then we knew then, you know what you gotta do, right.

So fast forward before I left DARPA I got to Christen the first ever self-driving ship. See Hunter that we put in the water. And at that christening ceremony, by that time, we were paired up with the Navy and the Navy was a partner with us for awhile. And I think now is taking the effort [00:20:00] forward.

And you know, now we have a working prototype. Now the Navy can say, Oh, let me figure out. Do I want to use it to hunt sea mines? Is it a cheaper, safer way to trail quiet diesel submarines? You know, there's a lot more that has to happen to really figure out how you take this and move it forward. So that's a success story, but I think that stealth is another great example.

These things were not only not embraced or asked for, or, or. Welcomed when they weren't delivered from DARPA, they were, you know, they were spat upon often. But it doesn't matter because if it's radically better enough and, and the stars align and you get like, I mean, a lot of things she can't control, but that is how big changes happen.

And you have to be able to do those things, even when there isn't a customer standing there waiting for it. I appreciate that. And so, yeah. Do so, so like then let's how does that then translate for you guys for actually, yeah, so I think the way to think [00:21:00] about it for any, any, so look, I mean, anytime you're setting out to make.

To spark a radical transformation. You, it's not going to happen unless you really think about the entire system of what it's going to take to, to create the change that you want to see in the world. And so let me just take one really specific example. One of our programs, Dave safes at actuate. But one of these is one of Wade's programs that he's built.

The objective there is to use privacy technologies that are emerging, that are currently being used ad hoc to build a new architecture and infrastructure that would allow for multiple data sets to be provided on an encrypted basis. And then what would allow researchers or policymakers, anyone who wants to analyze the data and cross-link among those data sets for the insights that they hold would allow them to do that entire process while rigorously preserving privacy.

And that includes. The CR the linking, the cleaning and the [00:22:00] linking, you know, all the sort of, or ugly data science stuff that has to happen before you can actually start seeing the insights. So it's a soup to nuts full system. That's the ambition of that program is to demonstrate something that's that's that's.

Robust enough and flexible enough to handle many different kinds of data and data problems. So the future that we want to see is that instead of today, where research is, you know, you ended up doing research or policy in halafu, it's sort of a lamppost problem, right? You do a lot of interesting research with the data you happen to be able to get a hold of, or that you happen to have permission to link to other data, but all the really interesting problems what, what.

Happens in K through 12, but that leads to different kinds of life outcomes. How has that to other environmental factors in a kid's neighborhood or the way that, that education and that child is going to end up interacting with the criminal justice system? How, how do all of those things tie to the progress of the [00:23:00] economy and jobs and the things that lift people up and allow them to pursue opportunity?

That's you know, to answer those kinds of questions, you need 53 different agencies at state local and federal levels, and you need private company data. And you know, like it's all just it's it exists, but that doesn't mean you actually can get at it and start using it. So we want to see a future where you could answer those kinds of questions.

Well, so what's it going to take the piece that the program will do when we're able to get it going is to demonstrate a prototype system that allows for radically different kinds of data owners to put their data together, you know, run some real examples and. And do applications show that are demonstrations of what this new data capability would look like, but that's probably not going to be enough.

Right. And so the other things that need to happen you know, my dream is there's a future where there's NIST or other standard for the kinds of. [00:24:00] Procedures and processes that would allow the legal counsel of the firm or the organization that owns the data to say, okay, if we comply with this regulation, if we meet this certification, I can now sign off and know that I'm protecting the data properly, but I can, I can make that decision tomorrow, not in six months or a year, like it usually takes today.

And, and, and then over time with, you know, with a lot of different players and. An infrastructure for regulation and certification, you can start to see how you could, you could have the kind of rich data future that, you know, w we all talk about these days, but actually isn't quite happening yet. So, so I think that, I don't know if that's a useful, for example, but what the pic, the general picture is.

Think about all the entities, all the actors that are going to have to. To do something to change their minds, take an action. And you may not be, I mean, we're not going to go fund all of that. We're going to fund a piece that would allow them to change their minds. And that's really, our [00:25:00] objective is a prototype and demonstrations that cause them to say, okay, we can, we can now do something in a different way.

Do you see encouraging them to change their minds as part of the program in that there's sort of like a very there's there's a spectrum of from just like demonstrating the prototype and then washing your hands of it too. Like. Push like knocking on their doors for years. And I assume it's somewhere in the middle.

Yeah. There's a lot of leading horses to water recognizing that you can't make them drink. What I, what I think is really clear for many, many years of experience at DARPA and other places is that if you're not deliberate and thoughtful about. Who those players are, what would cause them to change their minds and then doing the active work to engage them all along the process.

For sure. If you don't do those things, the chances are pretty, pretty slim. If you do them, you might have a shot. Right. And [00:26:00] and so I think we're as we're designing programs that actually we're being. Very explicit about that engagement process, which starts by you have a lot of conversations with people who are like, most often, they're like, yeah, sure.

You're in fantasy land. If that stuff existed, it'd be awesome. I'm like, that's not the reality. And let me tell you what I really need. So that's at the beginning. And then as a program starts, you know, during the execution of a program, that's really when it starts going from. Just, you know, something that the program leader believes in to something that now is starting to be palpably real potentially.

Right. And so you want to bring those. Decision makers whose minds need to be changed, but at least could be investors. They could be entrepreneurs. They could be policymakers. I mean, a whole different sets of who those, those, those adopters need to be the ones that are going to take it to scale. But the places where we can bring them to the table are you know, you continue to call them up and tell them what's going on.

But. But you [00:27:00] create demonstrations and updates where you bring them to the technology or you bring the technology to them and you say, look, did you, did, you know, this was possible. Look what we can now do. And, you know, ideally they get dazzled and then they say, Oh yeah, but they hear the next three things.

That would be a problem. And that tells you what you need for the next phase. So that's what, that's a parallel track to the three to five years of technical work that's going on in the program. That makes a lot of sense. And in terms of the technical work, do you plan on having it be mostly externalized to the organization?

The same way that DARPA does. I would say w there there's a very important piece of intellectual work and management and leadership that happens with the program leader and that individuals tiny little team within actuate, very much like at DARPA. But you know, the vast majority, the overwhelming amount of the funding goes out to the, the companies, the [00:28:00] universities, the nonprofits who are doing the different components of R and D and.

Testing and demonstrations and all the people who are doing all of that work. And that's for a couple of reasons. Number one you know, these are three to five year projects programs, and we w w what we want to do is we don't want to hire them all and put them under our roof for that period of time just as a practical matter.

But the other really important thing is when the program is over. What you want is, you know, a successful program and w a program starts with a program leader who has this vision. Yeah, they are, they are, you know, they're calling people to try to do this really difficult, new thing, and. At the end of a program, what you want is that entire community that you've been funding and working with that, they get the vision.

Not only that they built, they delivered it, right? Like they've actually built this thing and they become the most important [00:29:00] vectors for moving it out into the world and getting it. Actually implemented. So the world starts changing. And so for both of those reasons up front and at the back end I think that's, I think that's one of the powers of the DARPA model is, is tapping these amazing talents wherever they are.

Yeah. So something that I've actually wondered about with the DARPA model, that I've never been able to find any good information on is what do you do when you run into a situation where You need, like there there's multiple groups that have been working on different pieces and there's like, is there ever contention over, who's going to take it forward or like, like, how do you, how do you sort of coordinate it so that the outcome is the best for the world where like, which might involve like like squashing someone's ego or something like that.

I was like, shocked. I'm shocked. So are you thinking I would say they're somewhat different answers if those junctures happen [00:30:00] during a program versus after a program. So, you know, let's say you have a program that that had different university groups working on dunno some advanced chip for doing machine learning or whatever.

And, and, and it, I mean, this just happened. I think that there were multiple very good research results, but then were commercialized in different ways by the performers. So at that point, you know, it's like, great. Let them drive it out. Hopefully they, they. But they may compete with each other. They might go after different market segments, but there, there are multiple shots on goal to commercialize something coming out of a program.

And I would characterize that as something that DARPA would not particularly, I certainly wouldn't control, probably doesn't even have much influence. Conversely, if you're in a program at the early stages of a program, a lot of the that's a lot of what the core management Work is for the program manager at DARPA or the program leader as we're calling them at actuate [00:31:00] is, you know, so let's back up.

Number one, you're trying to do something that achieves huge impact sad, but true that involves taking risks because all the low-risk things have already been done. And so the, the whole art of this. Business is how do you intelligently take and then manage and drive down and eliminate risks. And one of the, one of the really effective tools in the toolkit for managing risk is a to S to S to plant a number of different seeds.

And to deliberately have competitive efforts that might, you know one of our programs at actuate, for example is built on the idea that we have all kinds of research that could be better at real-time incentives to help people make better. To develop healthier habits. So, you know, it, when we get that program going, we're going to deliberately have multiple teams who are working on different kinds of incentives, themes, and then a core [00:32:00] management challenge in a program like that is going to be, you know, you, you may choose to start four, but you, you know, at some point you're, you're going to want to down select and go to two.

And what is the right point? When is it. Point where you want to say, you know, I'm going to put more of my eggs in these baskets. And so I think that that's integral to the design and then the, the day to day or week to week management of the program. And I imagine that there might be one more situation where at the look you're actually sort of building a system and you have different groups working on different pieces of like different components in the system.

And so. What, what, how do you, how do you manage that at the end? Where it's like, okay, like at the end of the day we, we want the system.

Yeah. That's exactly right. Yeah. And I, I let's say maybe just one small point at DARPA. DARPA's running 250 or 300 programs at any moment in time. Right? So full-blown huge agency [00:33:00] relative to the scale that we're starting at zero right now at actuary, but in the DARPA portfolio, you will find programs.

You know, the self-driving ship program was a systems development program, Gantt charts, milestones, boom, boom, boom. Right on the very other end of the spectrum might be a very much more research oriented program. That's highly exploratory. There's a new physical phenomenon that looks like it could be interesting down the road, but right now you just want to have vibrant research and people pursuing the question in lots of different ways.

So there, there are many, many models. Yeah. Somewhat in the middle is probably where is, is what I would characterize where actuate will start and what we're finding in the kinds of programs that we're exploring is over and over again. Here's the pattern there. Number one, there's a, there's a problem for which we think there's a radically better solution.

That's possible. The reason we think it's possible is because not because of one new research result, but because there are a handful of different research areas that are advancing in interesting ways. But they [00:34:00] haven't yet those advances have not yet really been applied to the right problem or critically to your point, integrated together into a system that can actually follow the problem.

They're just like threads or hopes. Right? Yeah. And so that becomes, I think this is a classic template. For solutions R and D program at DARPA or an actuate. So a great way to manage those kinds of programs is, should think in terms of different tracks of effort. And the first track is to advance the research itself.

So it's applied research where you're, you're building on these, these threads and nuggets, which you're really aiming at the specific new capability that your, that the programs. The program's goal is to demonstrate that, right? So track one is applied research. The second track is building prototypes and that's often that's a different kind of performer.

It's someone who can integrate the different pieces and you can, you know, you can imagine a process where every seat. Three or six months, there's a drop from applied research into building prototypes. Right. And so, [00:35:00] especially for software tools, this is like the classic way you would do it. So every three to six months to see what's coming out of applied research, that's baked enough to put it into the prototype.

And so that that's. That's becomes a very good way to flow things. That's tracks one and two track three is now you got to figure out if this stuff is doing anything. So then it's, it's testing, evaluation and working, you know, trying to show that it works for the application or applications that you're going after.

And while there are different tracks, they interact, right? Because as you're learning what works and as you take the integrated prototype, so an integrated prototype for. But tool to help individuals choose healthier habits throughout their days and their weeks. So it's going to integrate a whole host of these different advances that are coming from different areas of a lie, including incentives, as I mentioned before, but, you know, ideally every six months or so as the prototype strop to testing, you start getting real feedback about this, this combination of.

[00:36:00] Sensing and coaching and personalized incentive. Is it working or is it not working? Right. And then, then you go through these iteration loops. So I think that's So, yeah, I mean, I think what, what, so what the program looks like when it's underway is you'll see some researchers, universities, or companies you'll see prototype developers, typically more companies there you'll see people who do the tests or the demonstrations.

It could be a clinical trial. If it's health-related it could be, I mean, it could be whatever, whatever the form of the prototype or the application is. And then throughout the whole thing, and the management challenge is. You know, you have a plan and then reality is going to happen. It's going to be something different.

So how do you keep that whole engine moving forward? That is, that is an amazing description. I really appreciate you going into those details. Cause I think that that's something that. People don't think about it enough is, is sort of like how, how to manage those tracks. I want to actually go back to something that you said earlier, which is that the people that you want sort of as [00:37:00] performers in the program are the people who can see where the boxes and then, and think about, think outside of it.

And do you have any, any strategies for finding those people and, and sort of teasing that out of them? Yeah, I, I think I said it more in the context of program leaders. And then, and, you know, by the way, at DARPA, one of the best ways to go find great new program managers or potentially great new program managers.

Cause you don't really know until you give them a shot. Is to find, go through the performer base. Right? And there, there, there at DARPA I found there were always, there were always performers who were very, very good at their piece of it and they loved their piece of it. And you have to have those people, but then once in a while, you'd see a performer who started seeing the whole picture and they could help the, you know, they would start being creative about like, we could go here.

And when you start seeing that, those are the, those are the signs. So I have a set of. Criteria that I thought about in terms of [00:38:00] DARPA program managers. And it's very similar for Dar for, for actually future program leaders. Number one, it's people who are driven to make a change in the world which like, I mean, this is where I live and breathe, but it.

Over time. It has finally dawned on me that not everyone gets out of bed in the morning to make the future a better place. All right. Like that's just like what the culture and the whole point of the exercises. They have to find people who are driven to do that. I'm always looking for domain expertise because you need to be deeply rooted and deeply smart about something that's relevant to the problem it's going to work on almost by definition.

You won't be a domain expert on everything that it takes, because these are big systems complex. Thanks. So the next thing I'm always looking for is the ability to understand the whole, the big picture of the system, and then to navigate seamlessly, you know, from, from forest to trees, to bark, to cells, right.

And then back up and you have to be able to do that whole thing. And that means you may know a little, a lot [00:39:00] about how you know how some aspect of behavioral science works in a very specific context, but you also, I'm also looking for people who can then extrapolate up to how might that and other advances to be harnessed, to, to move the world forward.

Right. And that that's that's I would tell you that's one of the harvest characteristics to find, cause of course. W w you know, there are lots of people who have domain expertise, but that ability to navigate from systems to details is, is actually a very precious commodity that I always love when I find I'm looking for people who, the overall thing I'm looking for is people who have, you know, head in the clouds feet on the ground, because you need to be able to dream, but you actually have to be able to go execute.

And in this case, execute by managing other people on projects. Yeah. You know, it's not an individual contributor role. And then the final thing that matters deeply is an ethical core, just because you know, that that's important for how you treat people on [00:40:00] a day-to-day basis. But it's also important because we're talking about really powerful technologies and someone who we need people who are willing to be explicit and thoughtful about the ethical considerations that they'll be weighing in.

Yeah. That that's great. I want to change gears just a little bit and sort of talk and talk about money for a little bit. So, so, so you spent many years in venture capital, and so I assume you, you know, the, the, sort of both the upsides and the downsides of, of startups and for capital organizations and you decided to, to start as a nonprofit.

And so, so I'd love to sort of understand the thought process behind that because I definitely, I, there, there's sort of a line of thinking that. You know, it's like, if it, like, if it can be done, it should be done as a company, as a, like a startup. And so I'm interested in why you, so I would say that [00:41:00] simple minded and, and to the extent you think that's, if that's your worldview, I would say the things I think need to be done, that I can make a contribution to cannot aren't companies.

They're not there. There's not a visible market. And so it's not, it's not a company today. Some of the things we want to work on will part of getting them out to the world will involve markets and therefore companies, including startups, but you know, coming back to these major societal challenges that we have none of them are simply going to be solved.

By companies, building new products, services, and profits. And I do think that some of the solutions will ultimately will include companies having really interesting new market opportunities. But it, you know, this is the stuff that the market doesn't do and, and. But, you know, th the, so if you think about us, R and D we spend about half a trillion dollars a year in the U S economy on research and development. [00:42:00]

The majority of that of course, is companies doing product development and but about a hundred, I think it's about 140 billion a year that's that's federally funded, R and D and and, and the, the. But areas in which actually is focusing are places where they are not market driven opportunities and, and they are not, I think they are not yet the places where we have the federal R and D machinery and yeah.

But so those things need to happen for our ultimate dreams to come true. Right. Is to make the difference that we want more. And, and ideally it seems like you, you'd almost sort of like pull both of those both of those leavers, like towards a certain direction, right? Like that's, that seems like a, a place that you could sit getting opportunities for them.

Right. I think that's the biggest pull as you show them something that, that changes their minds. Yeah. And are you funding the organization as like actually as an organization, as a whole? Or are you funding [00:43:00] each sort of program? Like, are you funding it as a program by program basis? We're still at a seed stage just to be really clear, but we spent a lot of time on this strategic question about whether first of all, let's be really clear that what we're trying, we think philanthropy has an important role to play because of the fact that market and government are not.

For various reasons, stepping up to the plate on these topics that said that what we're trying to do in the social sector is there isn't a template for it. It's not what philanthropy has, has done at least in the last, you know, Six or eight decades. Very interesting stories about Rockefeller foundation and the green revolution and how they, how they funded the research.

But, you know, if you go back and read how they thought about it in the methodologies that they developed, it looks a lot like solutions, R and D and then those. Actually those human beings, those exact people went into whenever bushes organizations on during the second war. And, and [00:44:00] I mean, that's the template for solutions R and D is right.

We have an existential crisis and we have things we can do about it. And it's all hands on deck and integrating everything. And. Building radar on the bomb. Right? So, so anyway, so, but it's been decades since part of philanthropy, I would say, was really seriously focused on this kind of solutions, R and D.

So with that, that is the significant caveat. So everything we're doing is going to be a big experiment in the social sector question you're to get now to get to your question. That we spend a lot of time thinking about whether we should try to build a program, build a program, go raise money for it.

Or if we should try to do something that's even harder, which is to raise a fund, to do multiple programs and build a portfolio we've settled on the ladder. And the reason for that is simply that, first of all, I think, you know, sometimes doing an impossible thing. It's better to do the more impossible thing that actually.

Can make an impact. I think this comes back to risk management and we talked about risk management within [00:45:00] a program, but a lot of, you know, how to start have one or two things, every single decade that literally is changing the world. Well, it certainly isn't because all the programs succeed, it is because you have a portfolio.

And because it's a very deliberately managed diversified portfolio, it's diverse in. Aspects of national security it's that it's targeting, it's diverse in the technological levers that it's pursuing, it's diverse and timeframes to impact. And so at the end of the day, we concluded that for actually to make a dent on any of these met massive societal challenges that we needed to be able to build portfolio.

Yeah, no, that makes a lot of sense. And so teaching to do tracks again and just talk to you a little bit about, about the Pat, like your, your, your, your career, which has included some like amazing things. Like when, when you became the DARPA director, like how, [00:46:00] how did. You know what to do? Like did they,

I'm sorry, this is a silly question, but as you say, it seems like such a big role. Yeah. I've been super lucky in the things that I got to do. But I th the luckiest day, I would say in my professional life was the day that Dick Reynolds, who ran the defense sciences office at DARPA in the 1980s.

He said to me at a workshop of there that I happened to be attending. He asked if I wanted to come to Darko as a program manager, and I was 27. I had been out of graduate school for. A year. Oh, maybe I was 26 at the time. Anyway, I had only been out of graduate school for about a year. And I was in Washington on a congressional fellowship at that time because I had decided I wanted to do something other than research on the academic track, but I didn't know what that was.

It's like on a [00:47:00] Lark, I went to Washington for a year, which was critical because even when you leave the trot, you know, th the, the path you are supposed to be on, that's when you don't know what's going to happen. But one of the things that can happen as amazing new possibilities occur. And that's what happened when Dick asked if I wanted to come to DARPA.

So at a very early stage of my career, I landed at DARPA and it was the first place I had ever been. I mean, I had worked. Two summers at bell labs who put me through graduate school. I'd worked at Lawrence Livermore one summer as a summer student. I'd worked at Texas tech and the laser lab as an undergraduate.

I'd done this graduate work at Caltech and then I'd been on at the office of technology assessment. And the honest congressional fellowship I got to DARPA and all of a sudden, it just made sense to me, right? Like everything that I thought and believed in the way I was culturally oriented, which was you go find really hard problems.

And then the contribution we get to make as technologists is we get to come up with a better way to solve a really hard problem. And we get to [00:48:00] blow open these doors to new opportunities. I just, it just resonated so deeply. So I spent seven years at DARPA the last couple of years, which we're starting with micro at that time, it was the micro electronics technology office, which we spun out of the rail defense sciences office at that time.

And I, I, you know, I loved it. It was, it was a crazy ride. Right. I got to do all kinds of things that were very, very meaningful then. And that, you know, for the 30 years, since then, it's been just. Such a delight to see so many things, but have come into the world that trace back to some of the early investments that we got to make.

And I would tell you that while I loved it, everything else I got to do after DARPA and I treasure it and I needed those experiences, I never really got over being at DARPA. It was just like, it was my home. It was my place. It was what made sense to me. And So when I got the call in 2012 to go back and lead it I, you know, it was just a dream come true.

And [00:49:00] when I got there, it was, you know, being a program manager and then being an office director at DARPA, which I had done in the eighties and nineties, and then going back as director, those are three very different jobs, but so there was a huge amount of learning and growth in every stage. But they are all.

Lined up to this mission and vision of an organization. That's just like, I'm wired the way that DARPA is wired. So, so I, I have to say it's, it was the most satisfying job that I've had so far, I'm trying to make actually even more. It was very hard. It was very meaningful, but I have to tell you, it just felt natural.

It felt instinctive and natural in a way that none of my other jobs really did. I have to say, I mean, you know, And they were all. Okay. And I think there are other jobs. I think I was good at their other jobs. I was horrible at that matter, but DARPA was the place where it just sort of, it just felt natural to me.

Yeah. And, and so sort of to provide on that and, and in closing do you [00:50:00] think that there are any ways to improve on the DARPA model that you're trying to implement going forward? So we talk about this all the time. I mean, I think for small, if the work that we're starting an actuator can have anything like the kind of impact that DARPA has had in, and, and, you know, any subset of its programs Then I can die happy, right?

Like if we can really make a contribution to these big societal problems, that's, that's, that's going to make, that's just going to be deeply meaningful to me. We've talked about some of the things that I think are difficult in the DARPA model. One of them is about the more radical the innovation and advanced the harder typically is to get anyone to.

Change the way that they work in order to adopt it and get the benefits of it. So I think being we're, we're trying to be even more deliberate about how would you get decision makers to change their minds and implement in the design of our programs have actually, I mean, I think DARPA does that, but that's something we're trying to put [00:51:00] special focus on.

I think DARPA's done a huge amount of work to make it easier to, they have legislative authorities and good practices about being able to hire people who. Many of them normally wouldn't consider public service for many reasons, but especially of course, low compensation levels. And while dark was not fully market competitive, we w we were able to move very quickly and had a little bit of a salary cap relief.

So, you know, the nonprofit sector is not going to be the place that you make your billions obviously. But I think being outside of government has that advantage and something that we'll, we'll definitely take advantage of. And they're, you know, they're things that are simply not appropriate for the government in a market economy.

To do. And so there, there are things that you can do for national security, but that, that unless we have a radical change in our thoughts about industrial policy, which by the way might be happening, I can't quite tell, but there are ways in which government has not [00:52:00] chosen in the past to work with industry or with finance that I think are less, you know, those are not as significant on a limitation for the work we're doing in the social sector.

Nice. Excellent. Well, I want to be really respectful of your time. How can, how can people find out more about what you're doing? And like if they, if they think this is interesting, like what, what should they do to, to help out. Well, thanks so much for talking about this. I love the fact that you, that you care about these issues and you've done more than anyone I've seen from outside DARPA to really understand the agency.

So that it's been so much fun talking with you, Ben, about that. I think you're going to provide the link to the issues in science and technology. And our website courses, if it's all brand new. So take a look and you know, we're so early right now, but I'm, I'm always looking for people who have a deep passion for these societal challenges who see new opportunities to do things that are radically better way.

[00:53:00] And please reach out to us from our website. If you, if, if it resonates, we'd love to hear from you.

Thanks for listening. We're always looking to improve. So we'd love feedback and suggestions. You can get in touch on Twitter at Ben underscore Reinhardt. If you found this podcast intriguing, don't forget to share and discuss it with your friends. Thank you.