Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.
During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.
Please contact [email protected] with any feedback.
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
In footnotes or endnotes please cite AIP interviews like this:
Interview of Arati Prabhakar by David Zierler on August 11, 2020,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
This is an interview with Arati Prabhakar, founder and CEO of Actuate, a nonprofit organization that aims to accelerate American research and development systems. Prabhakar recounts her family’s Indian heritage, and her mother’s decision to immigrate to the United States on her own and pursue a degree in social work. She describes her childhood in Lubbock, Texas and describes being the only student with an Indian background in high school. Prabhakar discusses her undergraduate education at Texas Tech in Lubbock where she majored in electrical engineering, and she describes the opportunities leading to her graduate work in applied physics at Caltech where she worked with Tom McGill on developing quaternary materials. She explains that her interests in real-life problem solving led to a fellowship with the Office of Technology Assessment in Congress, which in turn led to her government service at DARPA. Prabhakar describes her initial work at DARPA on gallium arsenide technology, and she explains the impact of the end of the Cold War on DARPA and on her career. She explains the circumstances leading to her move to NIST to lead the Institute where she focused on building up the Advanced Technology Program and the Manufacturing Extension Partnership. Prabhakar discusses the personal and professional reasons she decided to move to California to work at Raychem in 1997 and then Interval Research, and then Venture Capital, where she worked on funding semiconductor research. She describes her interests in clean energy and how she came back to Washington to head DARPA where there was a major focus on clean energy and pandemic preparedness. Prabhakar explains how and why DARPA operates in the realm of biological research and how she navigated the existential paradox of a leading an agency built on nimbleness within the world’s largest bureaucracy. At the end of the interview, Prabhakar explains how her career in both the private and public sectors prepared her for her current interests in utilizing research and development to confront macrosocial challenges.
Okay. This is David Zierler, oral historian for the American Institute of Physics. It is August 11th, 2020. I am so honored to be here with Doctor Arati Prabhakar. Arati, thank you so much for joining me today.
It's wonderful to be here. Thanks for doing this with me.
Okay, so to start, can you tell me... Obviously, you've had a lot of affiliations and titles over the course of your career. Tell me your current title and institutional affiliation? What you're doing now.
I'm the founder and CEO of a new nonprofit called Actuate.
And what is Actuate and what is its mission?
Actuate's mission is to accelerate a generational advance in the U.S. R&D and innovation system to enable it to be much better at addressing the critical societal problems that we face for the future. The problems that we're very good at the problems that Vannevar Bush identified in 1945. Actuate is about accelerating our ability to deal with today’s problems, like access to opportunity for every person. Like better health outcomes with lower healthcare costs. Being able to trust data and information, and being able to mitigate climate change.
Was this something that you were inspired to do knowing that these problems were only going to get worse with the incoming Trump administration?
No, that wasn’t the primary driver. I left DARPA in January of 2017, having had the great privilege of working in U.S. R&D and technology from very different vantage points. And my view was that we were very good at the agenda of the middle of the last century. And national security, economic growth, and jobs and health (which unfortunately got narrowed to biomedicine) none of those problems have gone away, but there are these other classes of problems that will now determine whether we succeed or fail as a society. It's a much broader issue than any one moment in time, or one administration. But I grew very concerned about these classes of problems. They matter but we don't know how we're going to solve them. We don't have robust innovation for those kinds of problems, and that's what we are setting out to try to build and accelerate.
And who were the most important institutional stakeholders that you're partnering with to achieve these goals?
Well, the point of the work is that we don't have the institutional capacity to do these things the way that we do for national security where we have labs and universities and companies and a defense department, etc. Or the health landscape, or these other things that we already have. Actuate is just at a seed stage. But our plan is to raise philanthropic funds to prototype and demonstrate what these new forms of innovation can look like. Because we think that's what philanthropy is for, to experiment with new things, try new things, and have a very diverse range of different approaches. And what works can be scaled, ultimately, by the public sector and the market. And that's how scale actually happens. But philanthropy, we think, is a good place to start.
And just to give a real-life goal, who would be an example of a beneficiary of all of the ideas that you have in mind, as they're becoming operationalized?
Well, the ultimate beneficiaries are individuals in our society. Especially in our very inequitable society. Really changing the calculus for people in our society who suffer very poor health outcomes and don't have ladders out, access to the kinds of opportunities that could allow them to break out of inter-generational poverty, for example. So in the ultimate sense, what we're really trying to do is experiments that lead to better policies and practices and in some cases products and services that change lives in a way that will allow our society to blossom rather than wither.
Awesome. Okay. Well, let's take it all the way back to the beginning, Arati. Let's start first with your parents. Tell me a little bit about your parents and their background.
My parents were married in India in 1950 just as the country was forming and standing up as an independent nation. They had an arranged marriage, and it was an unhappy marriage. They later divorced, something that was inconceivable for an arranged marriage in 1950 from India, but that would happen many, many years later. I have a particularly unusual mother. In the early 1960s, when she was in her early 30s and had a small child, she packed us all up and moved us to the United States. She was a social worker, and she wanted to pursue a degree that she did pursue and get: a masters in social services administration at the University of Chicago.
She did this by herself? She was already divorced at this point?
No, no, no, no. The divorce happened when I was in my teens. They got married in 1950, I was born in 1959. In the early 60s, my mom decided she was going to come do this thing in the United States. And because of some health issues, she actually came by herself, and then my father and I came a year later. He had a physics degree, I think, originally, and then studied electrical engineering. So you know, his background looked like every Indian and Indo-American of that generation, right? Technical guy, he came here. But the story of how we got to the United States is actually about my mom bringing us. And she came because she had worked as a social worker for a number of years, and wanted to learn social services administration, take it back to India. That part never happened.
Did your father want to go?
I don't think so. And you know, I mean, I think he ended up building a good life for himself over time in the U.S. Notably, when he retired many, many years later, he moved back home to be with his extended family and passed there. Late in his life. But the thing that really has shaped me and a lot of my thinking and my outlook on the world, I think really does trace to a woman in 1961 who would get on a plane and come to the United States and bring her family here.
Yeah. Did she know anyone here? Was there family, anything like that?
No. No, I mean the immigration laws changed shortly after that, but in that time, it was very rare for people from India to come to the U.S. at all in any field. And the way she tells the story is that she had this young child at home, that was me. She started exploring and she went to... I can't remember the name. There are these, or at least there used to be, these U.S. information service offices that just provided information about the United States. She wandered into one of those and started going through college catalogues and decided that that's what she needed to do, was go to the United States and learn how to run social services organizations. And she and her younger brother got on a plane in New Delhi. They flew to London, where he stayed to study engineering, and she flew on to New York and stayed at the Barbizon Hotel for Women, and then got on a train and went to Chicago. She took what she thought was the warmest clothing she could imagine, but of course it was inadequate to the Chicago winter. She was a vegetarian and learned to eat eggs, because in the early 60s in Chicago, you would probably starve if you were a pure vegetarian.
And she had great stories. Part of her work -- they had to do case work on the south side of Chicago, where the University of Chicago is, and one winter day -- she was freezing, she had to walk somewhere. She didn't have a car. And in that bitter cold, she apparently knocked on the window of a car that pulled up to a stop sign, and asked for a ride, and the really wonderful Black minister who gave her a ride to her destination, on the way explained to her why she should never, ever, ever do that again. She was just this 30-year-old woman from India who landed in Chicago and decided this was what she was going to go do.
Do you have any memories of India before coming to this country?
Very, very light ones. They're at the level of, when we moved to the Bay Area, it's the flora here, the smell like Punjab, at my grandparents' home where I lived when I was between the ages of two and a half and three and a half. That's about it.
Now, your mom obviously, it's, no matter what background she comes from, what she did is unique. It doesn't matter where she came from. But I am curious if her family background was such that international travel and higher education was something that was conceivable for a woman in that time? Or even if she came from a family of those means, as a woman, something like that still wouldn't have been conceivable?
She came from a family in which her mother was considered over-educated because she could read and write Hindi.
And my mom was the first one in her family, the first girl to go to school outside the home at all. And was supported in that, and then in going to college and everything, including coming to the U.S., by her father.
She may as well have gone to the moon, it sounds like.
Exactly. But, you know, she had a wonderful mother, but she had a father who enabled those kinds of huge dreams. And everyone in the family, you know, all three of her siblings ended up coming to the U.S., following her lead and with her sponsorship, and so that's the part of the family I know, is all the cousins and aunts and uncles on my mom's side. And she and they all point to her father as the person who made it possible. You just could see that he had the imagination and the belief in his kids to back them to do these things. It was sort of amazing to think about. A man born in 1896, I think, right? So isn't that amazing to think about?
And where did you spend your early childhood?
A few years in Chicago, and then we moved to Texas when I was 10 or 11, we moved to Lubbock, Texas. And I stayed there through my undergraduate years.
Why Texas? What was the opportunity there for your family?
That was a job for my father at Texas Instruments, which took us from Chicago to Dallas, and he got a PhD relatively late in life at SMU, Southern Methodist University, in Dallas, and then we moved to Lubbock because he was on the faculty at Texas Tech.
So where would you say, looking back, would you have spent your formative years growing up?
Uh-huh. And you went to public school there?
Yup. And to college, at Texas Tech.
Was it diverse at all? Were there any other--
Not at all.
No, no, no. I was the only Indian kid in my high school. People didn't know what Indians were. A man in the grocery store when I was a teenager said, "Where are you from?" And I said, "I'm Indian." And he said, "What tribe?" Without humor or irony. He knew Native Americans. He had never met anyone from India. So it was interesting, because there weren't enough of us for people to even have formed a view, leave aside have any negative feelings (laughs). We were just one-off, basically.
And, you know, it's a place I would never live again, but there's a real western spirit in a place like Lubbock. We had a dear friend whose grandmother had literally come to Lubbock in a covered wagon. And you know, it's a flat, dry, brown place where the wind blows all the time, and it cuts like a knife in the wintertime. So it was a sturdy, really, you almost say "brave" was part of the culture of that place. So that was a good thing--
Arati, would you say your family was fairly assimilated? Did they keep up any traditions or customs in the house?
We were very Indian in terms of food, in terms of religion, although it was a little bit more on the philosophy side, I'd say, rather than ritual driven.
Was there a Hindu temple in driving distance that you would be able to go to?
No, no, no. We were about the third or fifth Indian family in Lubbock, Texas when we moved there in, I think, 1970, if I remember my dates correctly. But we were, my mom raised me and my younger brother to think in terms of keeping the best from that culture and taking the best of American culture, and so it was, again, a lot of the attitude I have about that I think just traces right back to my mom.
I could guess pretty easily that your ambition and feeling that you could succeed at the highest levels, that might have come from your mom, but your sort of scientific sensibilities probably came from your dad? Would that be a fair guess?
Yeah, maybe. Sure, that seems reasonable enough. When I was a kid, my mom would start sentences with, "When you get your PhD..." And again, it wasn't funny, it was just like a matter of fact, and so...
Right, right. And when did you start to realize that you had these aptitudes in math and science?
I've always loved math -- I just thought it was beautiful. Wasn't that excited about science, which is incredibly boring usually in grade school. Or at least at that time it was. I am contrary enough and ornery enough that when people sort of indicated that that's not what girls do, it became more interesting to me. But it was fun, it was interesting. My mom claims that when I was about six I said I was weighing whether it was better to be a scientist or an engineer. (laughs) So, I mean, yeah, this is the gestalt we live in, if you're an Indo-American kid growing up in the United States. All the choices are open to you. You can be in any field of science, engineering, or medicine that you choose, right?
So it all seemed sort of natural. And it wasn't really until I got to college that I started, I realize now, that I started forming very clear views about science and engineering and what really mattered to me. That typically does come in college and early life, I think.
Did you think about going farther away for school, or did you know you wanted to stay close to home?
My parents were divorced when I was in high school, and it was not economically very feasible for me to leave Lubbock at that time. And I was very upset, not to be able to go away. I'd gotten into MIT and Rice, which were the only other places I applied, and I just thought the world had ended because I had to stay. But I got something out of staying at Texas Tech that really informed my world view. And it really was about the value of engineering and the core ethic of engineering. And so you don't know when you're a 17 year old kid, right? But it turned out that that's how life unfolded, and I got something really valuable out of it.
Meaning that had you gone to Rice or MIT, you might have pursued physics?
I don't know. I mean, MIT is much more—it also has a very strong engineering culture, so it's hard to say. You know, you don't know what would have happened. What I distinctly remember at Texas Tech was the electrical engineering department chair was preparing a bunch of young kids to get a bachelor's degree and go to work somewhere in the state of Texas. Most stayed in Lubbock, right? But almost everyone stayed in Texas. He was preparing them to go do a job and make a contribution to a company at the end of those four years, which is not what I did because I went off to graduate school. But that's what he was preparing people for, and what he told us was that everyone in a company has a role, you know, the people who do the financial management, and the accounting, and the people who do the marketing. But the role of the engineer is to create value in that company. And I just remember that that just lit me up. Like, we get to do that part, you know? And that was the beginning of really thinking about what I valued professionally. So that was really a good thing to get. I didn't get the physics and the chemistry and the math that I needed to go to Caltech as a graduate student. And that was a very tough transition. But I got something else.
With engineering, did you ever think about-- because there's much more of an industry culture with engineering, right? So did you consider at all not pursuing a graduate degree and going to work in the field?
No, not at that stage for sure. I was programmed form birth to go get a PhD.
Yeah, yeah. So how did you make that transition?
While I was getting my PhD at Caltech, I quickly realized that doing academic research, having that eureka moment, at two in the morning, was like the least interesting thing in the world-- like I was just dying because I hated it so much. And again, I learned some incredibly valuable things, but the most important thing I learned was that that academic research life and the outlook of science weren’t for me. You know, science's verbs are “know” and “understand”. Those are not my verbs. Yes, let's know and understand, but I want to do engineering's verbs, which are “solve” and “create”. And people who don't live in science and engineering see them as the same or a blur, and they're very interconnected, but the fundamental ethic is so completely different. And I love and admire science. I want it badly. But that's just not enough for me. When you know and understand something, that's a great starting point, but let's go actually solve a problem. Let's go create something new. That's what drove me.
And you know, I can articulate that now, 40 years later, but those views were formed by Texas Tech and its engineering culture, and its ethic about what the role of the engineer is, versus Caltech, even though I was in electrical engineering and applied physics, not in the, you know, high energy serious science part of Caltech. But the entire campus's ethic and culture is a physics ethics and culture, and it's “know” and “understand”. That is the temple at which they worship, and it turns out it's not my temple. So that was pretty good to learn early on. I couldn't have told you this story then, but that's really when it happened.
And what graduate schools were you thinking about attending?
MIT, Stanford, and Caltech.
And so obviously you had both the grades and the support and encouragement from your professors, that these kinds of schools were within your reach?
Yeah. I know I applied to Stanford and Caltech. I think I applied to MIT, but I actually don't remember now. But back then, you got in where you applied, so life was different.
And it was physics programs that you were applying to specifically?
Electrical engineering. I started in electrical engineering, I stayed in electrical engineering for my masters. The only reason I ended up having a degree that has the word "physics" in it is because when I got to Caltech, I had gotten some great advice from my beloved mentors at Bell Labs from the fellowship program I was in. And one thing they told me the summer before I went to graduate school is, go find a great thesis advisor. You're going to live with this person for five years. This is really important. And so you know, that turned out to be really important, because I immediately started hunting for a good thesis advisor, got close to working for one professor, and then heard some things from one of his students that I didn't love, and I ended up working for someone who had a foot in electrical engineering but was more of an applied physicist. And so that, you know, for all of that, just a happenstance of who I ended up wanting to work with. I ended up having-- my masters is EE, and the PhD is applied physics.
Now, in the physics world, of course, there's this big divide between experimenters and theorists, right?
Is there a similar kind of division in the world of electrical engineering? And I ask specifically because I'm curious what kind of work you thought you wanted to do when you got to Caltech.
Yeah. Experimental. That was always really clear. I'd had a couple summer jobs, and they were experimental in nature. One was at Lawrence Livermore, building lasers. Big gas lasers. And one was at Bell Labs, where they sponsored my PhD. I worked there the summer before and after my first year in graduate school. So experimental was definitely the direction I was going. I ended up joining a research group that was a blend of theorists and experimentalists and (laughs) my advisor was a theorist, and the--
Who was that? Who was your advisor?
Tom McGill. He was a wonderful solid-state physicist, applied physicist. And he's gone now. That was one of the smartest decisions I ever made. But when I joined the group, there were three guys in one of the offices next door to mine, and there was an old banged-up baseball bat in the corner. And I said, "What's the baseball bat about?" And they explained that it was handed down from senior experimentalist to senior experimentalist in the research group, and it was that person's job to keep Tom out of the lab, even if it meant using the baseball bat. (laughs) Because, you know Tom was one of those theorists-- Like Tom would come into the lab when I was running an experiment, and while he was talking to me, he would be turning a knob on a gas tank and screwing up my experiment, right? So we never actually used the baseball bat, but it was there for a real reason.
How did you go about developing your dissertation topic?
It started with work I'd done at Bell Labs. I was working on materials, quaternary materials, indium gallium arsenide phosphide, which was just starting to be used for lasers for transatlantic fiber optic systems, for communications. And there were some lifetime problems with the lasers and I got interested in trying to understand what was killing the lifetime of those lasers. And there were techniques to probe deep level traps in these semiconductor material systems. So I started with those materials, including samples that I got from my friends and colleagues at Bell Labs, and then added a different set of experiments that were looking at some lifetime issues in silicon that people were struggling with in more mainstream silicon production. We were moving to some new materials, gate materials, but finding lifetime issues. So I was digging into that.
When you defended, what were you thinking your next move would be? What kind of opportunities were you looking to pursue?
By the time I defended my thesis, it was really clear, at five years in, I very quickly I realized that I wasn't interested in the academic research track. I also-- I was a terrible graduate student. I barely passed some classes. I wasn't academically prepared for Caltech and I actually just didn't care. Like I didn't resonate with the whole thing.
But what about Bell Labs? Did something like Bell Labs resonate with you? That kind of environment?
No, not that much. I mean, that was still research. The thing that turned me on was when I could tell someone a research result that made a difference, right? And so that was always where I was headed. So the great thing about Tom McGill as a thesis advisor is, virtually everyone else on that faculty thought about a PhD as your passport to doing basic research for the rest of your life, and they couldn't conceive of-- and in fact, they talked about it with capital letters. If you thought about anything else, you were Leaving Research, capital L capital R. But my thesis advisor, his attitude was much more that this was a calling card to go do whatever you wanted to do, and even though he had some perspectives, but he didn't really know what else I could do. But he was open to it and very supportive, and just in brainstorming mode one day, he said, "You know what you should do is you should do a congressional fellowship." And neither of us had any idea what that meant. I mean now, the AAAS fellowships program, I think, is very widely known. It was still in its infancy at that time.
So he had just barely heard of it, and we knew of one or two ways to apply for that, and so I did that, and then one of my Bell Labs mentors who in a phone call I'm sure I told him this, he sent me a clipping from an IEEE newsletter. You know, like in the old days. He sent me a physical clipping from a physical newsletter (laughs) about a fellowship at OTA, the old Office of Technology Assessment in Congress. And I applied to that one as well, and that's the one I ended up getting. And I did it purely on a lark. All it was was a way to leave the known track, right? Like I was not going to stay on this known track but I did not know what I wanted to go do. I had no view, negative or positive, about public service, about the role of the federal government. I got nothing, right? I just, like I'm going to go do this on a lark. And it completely changed everything. It was great.
What was your next move?
I got the fellowship, I went to OTA. I did a study on microelectronics research and development. You know, a study that was meant for congressional staffers to read. And which gave me a calling card to go call up anyone I wanted and ask them a bunch of questions. And I remember I called up Gordon Moore, and he kindly agreed to see me in his office at Intel, and I had the great pleasure of interacting with him a few times. I don't know him well, but I got to interact with him over many other years. But that's the first time I ever met him. I got to sit in Gordon Moore's cubicle at Intel, circa 1984-85, and he said many wise things, but what I remember is he complained about the people at Intel who would leave and go start up companies, and I'm like this wet behind the ears kid and I'm sitting there thinking, "Wait a minute, wait a minute! Isn't that what you did?"
So it was a great experience. And in the course of the interviews, one thing I did was call up Dick Reynolds at DARPA, who had been the sponsor of some of the work that I had done in graduate school. I don't remember if I met him then. But through Tom McGill, my thesis advisor, it was kosher for me to contact Dick. Which I did and I remember I went over to his office at DARPA to interview him for my study, and we're sitting in his office in Arlington, Virginia. And it's a site that-- It's two DARPA buildings ago, very old history now. And I remember sitting-- he had this corner office and because it's on the flight path going into National airport, while I was sitting there in his office, this plane almost feels like it's going by and you could almost reach out and touch the wing. But in a good way. This is before 9/11, so it was still like an awesome, cool thing going on. And it was like this magic feeling, because I got to come talk to Dick Reynolds.
Well, a few months later, Dick had invited me to a workshop. This is a classic thing at DARPA - you're running a program, you've got, I don't know, five or ten or 15 different companies and universities and labs that are working on different parts, and so at least annually you get them all together and you have a big program review, and everyone talks about what they're doing, and Dick invited me to one of those. And during the break, he asked if I would think about coming to DARPA as a program manager. What he said was, would I come to DARPA? And I was like, "Wait, wait. I know I'm not qualified or experienced enough to be a program manager, and I don't know what-- Like there aren't any other jobs there." It's like, what is he even talking about?
And it turned out he was crazy enough to take a chance on me as a program manager. I was 27 years old, I was a year out of my graduate schoolwork, and I think that was-- I mean, as far as I can tell, it was the fact that I left the standard track, right? Like that's a statement to the world that I'm going to do something else, and I'm thinking about the world in a different way. And I think that's it.
But what would be the standard track?
Oh, a tenure track position at a university or, in my time, the only standard path was tenure track position at a university, or you went to Bell Labs or you went to IBM Research. That was pretty much it for semiconductor researchers. That was what you did.
So just to understand, the fellowship with OTA, up until that point, was not off the track? This was not an off the wall thing for--
Oh no, that was-- No, no, no, that was completely crazy. In the context of the track I had been on, the PhD, go do research.
OTA was crazy. People thought I'd lost my marbles. It was already off the tracks.
No, so I mean, my question then is, once you're already off the track, how is DARPA... I mean, haven't you just sort of replaced the concept of what the track is at this point already?
Yes. Yeah. I mean, I think when you do a fellowship, the charm of a fellowship to me is, it's a chance to try something new and I think at that point, I could have still gone back into the track.
But I think my point, all I was really trying to say was, you know, why would Dick have even thought to ask me about coming to DARPA as a program manager at that stage? I don't think that would have happened if I had been on the track. And I think it was the fact that I left the track that made it possible for me to have that chance to go to DARPA at that stage.
I'm curious if, you know, you've already explained well how you were sort of not interested in pursuing academia, which you sort of intuitively understood even as a graduate student. And even with Bell Labs, you were not so much into the research.
So I'm wondering if you were primed? Not that you knew what sort of political life in D.C. was like, but if you were sort of intellectually primed to accept a new kind of future for yourself, whatever you might have found in D.C.?
I was young, I was unencumbered, and it's a pretty easy time to try something. And I did not move to Washington thinking that I was going to stay there.
But you know, when you're... I got my PhD when I was 25. At 25 you can go do that, right? I moved across country with all my worldly belongings not just in my car, but in the trunk. This little, tiny trunk of my car. So I think you just have flexibility, and I don't remember feeling... I'm sure I felt anxious, but I don't really remember feeling anxious. Like it was clear that I was employable and you know, I was going to go do this.
The other thing I thought about doing, I got an offer to go work-- I did get offered to go work at Bell Labs. There wasn't the research part of Bell Labs that had sponsored my fellowship, but it was some other part of Bell Labs. And I had interviewed for other jobs when I was leaving Caltech as well, including places like Intel. So it didn't feel risky. It just felt like I was going to try a few things, and before I really had to decide, the DARPA thing came along. And you know, as I thought about it, I thought that was the right thing, so.
To what extent did OTA give you a really valuable, broad view of science in the federal government? At this young age, was it an opportunity to really figure out some big questions about science policy in D.C. generally?
Mm... It wasn't that much. I was 25, 26, and 27. So I was a baby. And I was working on one particular thing. And the picture you're describing is something that takes a long time to really wrap your head around, I think. But I certainly saw... What I wanted was a life in which I wasn't sitting in the lab by myself going deep and narrow. What I wanted to explore was a life in which I saw a bigger picture, and for sure I started seeing that. It was definitely the beginnings.
I'm curious if also you started to see yourself as a policy person with a science background?
No. I still don't even think of myself that way. Because you know, running programs at a place like DARPA, you're implementing policy, but the role-- I always thought of my role in the context of the R&D and innovation system. And if it was bench research, that's a particular role. And OTA was a fellowship and I stepped out of that. It was more observational. In fact, one of the things I didn't-- one of the reasons it was clear pretty quickly that I didn't want to stay at a place like OTA is they are in the "on the one hand, on the other hand" business. Their job is to reflect both sides and not draw conclusions, but serve it up to Congress for their drawing of conclusions. And I quickly realized that I loved that, I learned a lot, and I honored that, because I think it's valuable, but I wanted to pick a direction and go, right?
And so when I got to DARPA, number one, I felt that my role had changed. I think of that as, now you're making a different contribution to the R&D and innovation enterprise. And for me, what happened is I got there, and all of a sudden it was like boom. This is what I was born to do. The way they thought about things, the way I thought about things. It just felt deeply worthwhile and I was good at it. I wasn't actually that great at doing basic research or taking classes at Caltech. Ugh. When I got to DARPA, it was great. It was just fantastic.
Arati, you said, I mean your immediate reception to this was, you're young, you're not qualified for this, you don't understand any of these things. But I wonder once you got there, looking around at some of your other colleagues, how out of place you were, you know, with a PhD from Caltech?
I was fine. Dick Reynolds hired me, my immediate boss and colleague was a man named Sven Roosild, from whom I learned everything about how you operationalize DARPA programs. Everything from how do you move money, how do you make selections, etc. To the really-- the ineffable things. I remember I said to him, "Why will this guy do this thing that we want him to do?" And he explained what's in it for everybody, right? So the gestalt. He gave me all of that. But you know it just immediately clicked, and so I think very quickly, I was able to be good at that job, and most of my colleagues were 15 or 20 years older than I was, and they had a huge amount of experience, but a lot of this is about a way of thinking. And it just clicked, and I was really good at it. I mean, in your professional life, that's what you're always looking for. Is the place where it's aligned, and you can just go. And I hit that. That was the first time I really hit it.
So what were you doing? What were some of your first assignments at DARPA?
Gallium arsenide technology. Gallium arsenide is now used quite widely. There's some in every cell phone to transmit a radio signal. But at that time, DARPA was chasing building digital circuits out of gallium arsenide, partly because of the radiation hardness of gallium arsenide. It turned out that it was how you build high speed, high performance circuits and the defense part was it was radiation hard, good for space applications. But of course, over time digital silicon completely ate digital gallium arsenide. Like it burped, and that's done, decades ago. It was a classic thing. We worked really hard, we made some progress, we did things that no one thought was possible. The thing we thought was important to do got completely over-run in the course of time. But meanwhile, a lot of the materials work that we did ended up being the seed bed for what did work, which was more on the analog and microwave devices and all of that.
Did it feel like you were working in a national security/military environment at DARPA?
Sure. You are, and it does feel like that.
In what ways? Like, on a day-to-day? Are you seeing uniforms, are you talking about the national defense?
I had to get a security clearance. I didn't see a lot of uniforms because there are, I don't know, when I was director, it was maybe 8 or 9 or 10% of the program managers were military officers in uniform. I don't remember what it was when I was there early in my career. But they tended to be more on the military systems side of DARPA than the enabling technology side, and that's where I lived. So initially, yeah, I mean, you were very aware that you were part of the Defense Department. I mean, the goals overall were very defense-oriented goals. But the work was done with universities and companies that included defense contractors, but also included small start-ups and commercial companies. So that was quite a mix. And it sort of depended on what you were working on at DARPA, but it was of course, and obviously, part of national security.
Were the projects varied enough where sometimes it would feel like you were operating in a basic science environment, and in other times you were operating in very mission-specific environment?
Yep, absolutely. DARPA's portfolio spans those. When you come in to DARPA, a typical thing is you inherit the management of some programs and projects that are already underway. And then your job is to design new programs and win funding for them, and then run those programs. So the ones I inherited when I walked in the door were the more research end of the spectrum. Not surprisingly, given my background. And over time, I ran some of the more pilot-line manufacturing-oriented stuff.
But then I was starting my own programs. This is a theme with me: what I became concerned about was we had all this cool new stuff that was actually working, but it wasn't getting into military systems. So the program I crafted was actually the first time I did work directly with the military. I had to go talk to a bunch of people who were in uniform, trying to improve the performance of this sensor on this aircraft. And so yeah, so the whole span, even as a program manager, I saw that, and then in subsequent management roles, that you see that whole spectrum as well.
Now, the end of the Cold War sort of happened smack in the middle of your tenure at DARPA. I'm curious--
That was interesting.
I'm curious from your vantage point, how that changed your professional life?
Yeah. Two stories that really stand out. One was, we had a meeting of the technical office directors. It was the DARPA Tech Council of that time. And one of my fellow office directors who did nuclear monitoring, seismic monitoring, around the world, he came in with a viewgraph, because we still did viewgraphs, with a map of the world, and there was a mass of land that said, "USSR." And when he had made the slide the night before, there was one, and that morning there wasn't one. And that was really boggling. I mean, you just can't imagine being on this side of the Cold War ending.
How boggling it was that it ended. I started at DARPA in 1986, and it was inconceivable that the Cold War was going to end - and then it did. And then the other moment was classic. Another one of my fellow office directors ran the undersea warfare office, Charlie Stuart, and once in a while at that level, he got to brief a really senior person. Charlie got to brief the chairman of the joint chiefs of staff. His name was Colin Powell. And the way Charlie told the story was that he briefed his entire professional life, which is about submarines, and undersea warfare, and at the end the chairman of the joint chiefs of staff said to him, "But we don't care about submarines anymore." (both laugh)
So you know, it was definitely a pivot point in national security, and I think it took a long time. Arguably, the Department is still figuring it out, because it's a very complicated world. The Cold War was complicated, but we thought we understood the shape of it and the players in a particular way, and it's much more of a free-for-all now. So it took a long time. I left DARPA in 1993, and the department certainly had not figured out what this next era looked like, and there were all kinds of things bubbling at DARPA to try to figure out what this next era will look like.
Now, what was the impetus behind the Microelectronics Technology Office?
When I came to DARPA, I was hired into the Defense Sciences Office, which is the most research-oriented office at DARPA. Think of it as the bubbling cauldron -- looking at what's happening in research that might become the source of surprise that's really important to national security. But it's more exploratory in nature. I came in because within that bubbling pot, the microelectronics part, which had been almost nothing, had started growing very rapidly, including this gallium arsenide work, and it was up to $30 million a year. And it was sort of large enough that they wanted to bring a second person in to help run that.
In the course of my seven years at DARPA, my first tour at DARPA, microelectronics and electronics generally grew from about $30 million to I want to say about $700 million. And it was many, many different things. We talked about gallium arsenide. There were some programs in the Pentagon on silicon technology, on manufacturing technology, on display technology, that all came over. SEMATECH, the industry consortium that was 50/50 funded by the government and the semiconductor industry to strengthen semiconductor manufacturing technology, that started just shortly after I arrived at DARPA. And that was all about building a domestic manufacturing capacity because of its importance to national security. So all of these different things were bubbling up, and so when I came to DARPA, microelectronics was $30 million embedded in one office. By the time I left, it was the focus of actually two separate offices, one of which I ran, and one was electronic systems run by my friend and colleague Lance Glasser. So it was just a time when it was coming to fruition, and so much was going on that it was expanding significantly.
And you mean so much is going on, even beyond the national security world. Just in terms of telecommunications generally?
Well, first I would say the national security world cannot and does not live off in a separate area. These are so interwoven. You know, both in the spillover sense of DOD putting money into basic research that then ends up populating the commercial world. But also in terms of DOD's relationship with that commercial technology for its own needs. So it's very interwoven. And yes, it was absolutely a time in the broader community that the US semiconductor industry had been world-dominant. The Japanese were scaring everyone to death back then. Everyone was extrapolating Japan's rise, and something else happened. That's a good thing to remember as we're now extrapolating China's rise -- I mean, something big is happening, but it won't be a simple extrapolation again.
So all of those things were happening.
And were you at some point looking for new opportunities beyond DARPA? Or did the NIST opportunity sort of come out of the blue for you?
It definitely came out of the blue. There was a time when I was thinking about leaving. And you know, I interviewed for a few jobs, and didn't get excited enough to go. Within DARPA, I first got promoted to be deputy director of the office I was in. And then Vic Reis did a major reorganization, and he was the soul brave enough to make me an office director. Good for him. So I had three different jobs in the seven years I was at DARPA. But the call to go to NIST was literally a Saturday or Sunday morning phone call. It was the weekend after the Clinton inauguration in 1993. And I thought it was a prank call, so it was completely out of the blue.
And who was on the other end of that call?
A woman named Amy Bondurant, whom I didn't meet for many more decades, who was working White House personnel during that time.
And is this a political position?
Yeah, yeah. It was at the assistant secretary level at that time that I served in it, and it's later been bumped up to an undersecretary level.
Now, how old are you at this point? When you get this call?
I mean, isn't this a remarkably young age to be heading such a big organization?
Yeah, everyone thought it was crazy. You know, I looked like a poster child for ethnic and gender diversity from the new Clinton administration. I just looked like their worst nightmare, right? Like female, young, I don't think they cared about the Indo thing. And from the defense department, right? I looked like a nightmare.
(laughs) Had you had any contact with NIST? I know that NIST and DOD do partner up on some things. I'm curious if you had a sense of NIST and what went on there?
Just a little bit. I talked to them, I think I talked to them when I was still at OTA. And I think they even talked to me about a job at that time. So what I really knew about NIST was what they do in semiconductor patrology, because that was my particular field. But that was about it. And the reason I was being asked to come lead NIST was actually less about the Bureau of Standards part of it. In the late 80s with the name change, that was really about expanding the Bureau of Standards now to NIST to include not just this research enterprise to support measurement standards for the economy, which goes back to the Constitution, but also at that time, there were a trio of new programs that were started in the late 80s.
In 1993 when the Clinton administration came in, those three programs were at a fledgling stage. They had been piloted and-- you know, NIST was doing a good, earnest job of starting them with the tiny amounts of money that had been appropriated. And then the Clinton administration, I don't know if-- you don't look like you're old enough to remember this, but during the campaign, it was, "It's the economy, stupid." And as part of their economic focus, their science and technology strategy, their flagship initiative was to build up one of the NIST programs, the Advanced Technology Program, and-- actually a pair of them, that and the Manufacturing Extension Partnership. Again, they started in the late 80s, little bit of work, but the Clinton administration wanted to take them to full national scale. And that was the link to bringing a DARPA person in, right?
So there was some specific interest to bring a DARPA person in and then they decided you were the person, is how that happened?
Oh, I'm pretty sure this was-- I actually think I know, but I'm not going to talk about the conversation that led to this.
But it was not as directed as, "Let's go find a DARPA person." People were talking to people and getting advice, and someone said, "Arati." And that's how it happened.
So you're young, you have a basic understanding of NIST, but obviously not a deep understanding. How do you go about developing a plan of action, a mandate, for what you're going to do when you take over this incredibly large organization where you're going to be managing people who are almost entirely older than you? Probably everybody is older than you at that level.
Oh yeah. Yep, that's exactly what was going on. So the way this happens is, first they contact you from the White House and you go over and interview with the Secretary of Commerce, and then for a while you're a rumor. And some things happen during the time that you're a rumor, but the most important thing you're doing is not talking to anyone about anything. Then, the White House announces its intent to nominate you. That gets published in the newspaper, and then all of a sudden everybody knows that this is going to happen. But your job is still to not talk about anything. However, at that moment, you can start going over and talking to the people at NIST, who are like, "Oh God, what's about to happen?" Right?
So that's the stage at which I was contacted, I think, by... I don't remember the details. I suspect what happened is I was contacted by the person who was the deputy director at NIST, who would have called me up and said, "Do you want to come over and let's start talking about this?" And I remember I drove over in my car, and I parked and I went upstairs, and Ray Kammer had been at NIST maybe 25 years at that point. He was the deputy and knew where every skeleton was buried. And (laughs) he turned out to be a fantastic partner. I mean, we were just, we were a wonderful, wonderful team. But that first day, he walked me through the portrait gallery that has the oil paintings of all the former directors, and all the people who have won accolades for their work at NIST, and the Bureau of Standards. And he said-- I remember we were walking by the portraits, and Ray says, "You're the wrong gender, you're the wrong ethnicity, and you're the wrong age." And he said, "Only one of those is actually going to get better in time." (both laugh) It was classic. But yeah, he didn't say it with hostility, he was just like, we're all being really clear about where we are right here.
But I'll tell you what happened. Here's what happened. They had briefing books, but I don't learn by reading briefing books. What I learn by is I love talking to people and just pulling it out of their brains. So I had a bunch of those conversations, and very quickly, it dawned on me that I was about to go from one part of the executive branch that was highly functional and where people ran to work in the morning because they loved what they were doing, they believed in their mission, it really mattered. I was about to go to another place that had all the exact same characteristics. Radically different mission, very important, again, in this R&D innovation system, and in the functioning of our economy. Super important role. People who were deeply dedicated and this was what they lived for, was to do that mission. And when you've got a mission that matters and people who really care about doing that, then you can do great things, right? And so there was a mutual process of me falling in love with NIST, and learning what it was doing and what could be done better with the additional resources that we were now in line to get. That was sort of core NIST.
And then it was a whole adventure about how do you take these fledgling, new activities and really build them up? So that was more, pretty much clean sheet of paper. But you know, we bonded, and I really came to love that place, and because we were getting all this funding for the new programs, that became an opportunity to address some long-standing funding needs for the core-- You know, the National Bureau of Standards research labs part. And enough people won. Enough people saw improvements in the resources and the ability to do the things that they needed to do, that in the end I think it ended up fine. I know I loved it and I really respected the people there, and I think it ended up being mutual over time. So it was a very good experience.
At what point did you feel that you had built up enough credibility? You know, being all of the wrong X, Y, and Z things, right? At what point did you feel like you had built up enough credibility that you could just sort of lead and move forward to accomplish your goals without having those additional hurdles of proving yourself?
I never felt that way. I mean, the way you do any job is you come in and you figure out what's going to make it better. What's going to make it easier for people to accomplish the mission, and if you're mission-driven and you are listening to people and then integrating what they're saying, and using it to drive to the mission. You're going out and telling the world, starting with Congress and the Department of Commerce and the White House and then the whole outside world that NIST needs to deal with - I spent a lot of time getting out and explaining what we were doing. And when your people hear you telling stories and understanding what they're doing and celebrating it, and engaging the world and making your organization, getting it the recognition and the kudos that they know they've earned, then you're all just the same team. That's how you win credibility, I think, with your people.
So much of the physics world during the Clinton administration talks about the effect of the so-called "peace dividend," right? The ways that the end of the Cold War had really fundamentally reoriented budgets across the federal world. In what ways, if at all, did you feel those changes at NIST?
Well, I think the push in the Clinton administration to strengthen NIST, to strengthen its traditional almost 100-year-old core, and then to build these new civilian technology programs in partnership with the private sector, all of that work was a perfect example of this early 90s shift. It was a refocus and reinvigoration of the federal role working with the private sector on economic growth and jobs. And a slice of what was going on for sure had to do with the end of the Cold War, and turning to other matters that had perhaps not been as visible during the Cold War focus.
Arati, can you talk a little bit about the policy process in terms of who would you be most reliant on to figure out what's going on at NIST? At the level where, these are items that you need to communicate up the chain in the executive branch? Can you give a sense of where you are in that communication line?
Is this, when I'm underway as director?
Okay. So your question is, how would I internally figure out what I need to--
Do to deal with outside?
Well, the answer to that is a phenomenally capable staff. I had Ray Kammer as a deputy. He knew where all the skeletons were. I had very strong lab directors for each of the multiple labs at NIST. But we had a special office at NIST. It was called the Program Office. It was run by an amazing woman named Elaine Bunten-Mines. What she did was she rotated people from the bench level in the labs up to her office for a stint. I don't remember if it was months or a year or something.
And they basically became-- So like if I had a question about, I want to craft a story about... what would have been a thing of the time? Oh, okay. So we were thinking about, how do we within the NIST labs, what's the role for NIST in the information technology realm? It's a measurement and standards lab, back to the Constitution, and so we needed standard weights and measures, so we got that in a physical world, but what is the equivalent role for a standards laboratory, in the federal system, what's our equivalent role in the information world? So one thing you need to do is figure out what is it you're already doing, and then you need to probably have some conversations with people and bounce ideas off the walls, or off of each other, about how can we think about, describe, and then maybe change and advance the things that we're doing?
So Elaine's little office would be the people who would go dig information out or find the right people in our 3,000-person organization. They were phenomenal. And all those people were groomed for leadership roles, because they saw the big picture. They came up out of the lab and they saw a bigger picture of NIST, and then over time they all, many, many, many of them ended up leading different parts of the organization.
What did you see as your primary opportunities to capitalize on some strengths that you had inherited at NIST, and what opportunities did you see that there were significant issues that were solvable under the tenure, under the timeline that you knew you were going to be there?
So what the first part of the question was, say again? Was...?
Parts of NIST that were working well that you inherited but that you saw opportunity to take to the next level.
The traditional Bureau of Standards, the research labs for measurement standards, that was a 90-year-old culture. Facilities, people, who had been there for long careers. Deeply understood culture. So we had a great thing going. And an engagement with industry, a degree of respect and connection that was super important. It was working.
I added two things to that, I would say, with the additional resources that came in. One was that we were able to sharpen our mission focus. When you have a big laboratory and 3,000 people you have to feed, and funding has been flat and sort of withering a little bit for a lot of years, which was the situation at that time, what happens is you start hunting for other places to fund you and so you go knock on the door at defense labs or the Energy Department, or anywhere you can find who might fund you. And over time, that leads to a little bit of a mission drift. Because you become a little bit more of a shop for hire, rather than waking up in the morning to be the nation's measurement standards research lab. With the new resources directly to NIST, we were able to re-center on our core mission. You still want those ties, with all these other parts of the government. But you want the core direct funding for the mission, so that you don't have mission drift. And I think we got better at that, so that was good.
And then the second thing was what I just mentioned about information technology. The IT-related work in the NIST labs was mostly math and computing to support the other, the physical research labs, and we reconceptualized that and basically said, "We have an important role to play in measurement and evaluation standards for the information revolution that's going on and we need to serve that, not just inside NIST. A great example of that, by the way, that I saw while I was at DARPA, is for years and years and continuing to this day, as you're training artificial intelligence systems or evaluating them initially, and now of course you can train them as well, the data sets that you use, the standard data sets that you use for that, have always been the Bureau of Standards and now NIST. So that's a classic non-physical thing that's super important to the information world. So I think I helped a little bit, and I'm glad for the improvements we made.
And then the other three programs, the Advanced Technology Program, the Manufacturing Extension Partnership, and the Malcolm Baldrige National Quality Award, each of those had done something already before I joined. Each was on a very good track. But each one was really quite small, and we expanded ATP from... I won't have the exact numbers, but I think it had been funded at about $11 or 18 million per year before I got there. We ramped it to $300 million a year. And a lot of thought went into how to do that well and effectively. And some amazing things came out of that, so I'm super proud of that. Manufacturing Extension was how do you help small and medium size manufacturers modernize and use new technology so that they can be competitive on a more global economy? And that program, I think there had been seven centers established before I got there, and we took it to full national scale and had full national coverage working with all 50 states in that program. ATP eventually got killed. It died a political death in subsequent years, unfortunately. After doing some very good work. But the Manufacturing Extension Partnership is continuing to this day.
So in terms of legacies and accomplishments, you had the pleasure of both seeing an immediacy to some of your changes, because these happened during your tenure, but you've also been able to follow from afar, how certain policies and programs you put in place have continued to yield terrific results.
Yeah, that's the ultimate joy of anything we do in R&D, right? Certainly, both times I've left DARPA, and NIST as well, there's this long tail, and you hear a story and you think, "Oh, I know how that started. And I'm really glad we did that." That's a good feeling.
Were you thinking by 1997 that you wanted to go into private industry?
Yes. But there's a whole story about that. So, in 1993, I meet this guy that I had heard about for years before. He's the civilian technology policy staffer on the Senate Commerce Committee, and he's now become incredibly important in my life because he is handling my confirmation hearing. So that's how we first met. The punchline of the story is that this fellow, we're going to celebrate our 25th wedding anniversary soon and we now have two daughters. And so NIST was really important for some very, very critical personal reasons as well. By 1996, Pat and I had been married and we were expecting our first child, and by complete luck I had married someone who was from Northern California, from the Bay Area, and was thinking that this would be a good place to move back to. When I came to Washington 13 years earlier, I thought I was going to be there for a year, and over time I just imagined I would be in Silicon Valley next. Where else would a person like me end up, right? And so that was just luck, that we both wanted to come back to California. And so when I took maternity leave at the beginning of 1997, that baby was born on the day that Clinton was inaugurated for his second term. That maternity leave basically was when I left NIST and we moved back here.
You wanted to do this? I mean, it was just totally not feasible to continue on in that position, being a new mom?
No, it was just time. It was just time. I had never thought I'd be in DC that long. And you know, I had four years at NIST, of which the first two were a White House that backed us to the hilt. And a Democratic House and a Democratic Senate. And you know, everyone wanted us to take more money and go do great things, and they were just super-charging us. And we went and did all of that. In 1995, the Gingrich Congress came in, and so the second half of my four years was a White House who loved us and fought for us and defended us, including a veto threat on an Appropriations bill because of the cuts to our program, so God bless the White House. And a House that was just out to kill us. Every time I testified, they would literally be asking where it was in the Constitution that we had to do this stuff, and we were pointing to the part that talked about why you needed measurements and standards. And they were trying to kill the Department of Commerce and it was much more of fighting a defensive action. Which we needed to do and we did, but enough is enough. It was time to go.
Now, this was a maternity leave. You were just sort of not thinking about leaving the workforce, you were just thinking about whatever your next opportunity was, it would come about at the right time?
No, I took maternity leave and during that time, was starting to hunt for my new job. So we were thinking that that was the right time, that we would use that period to transition from DC out here, and that's essentially what happened.
And I'm just curious. Was your lack of interest in academia as strong as ever at that point? Did you think about like--
Nope, never thought about it.
Never thought about it? Even in an administrative role? Like as a dean of a college or something like that? Not your scene?
Nah, it's not my scene. I mean I love it, I've funded so much work at universities that I am so proud of that I'm just thrilled to be affiliated with. There's a professor at Stanford, and I ran into him at the Whole Foods right before the pandemic. Every time I've seen him over decades, what he does is he tells me how many PhD students graduated on the molecular beam epitaxy system, that my project from DARPA funded back in the mid-1980s. And it's up to 100s of students now.
That's so cool.
So we're in the Whole Foods, and he still tells me about this. It's just the best thing ever, right?
So love it, but it's not my scene. And boy, talk about a caste system. At universities, you're either a professor or you are other, and who wants to be other? So I've never imagined that. What I actually wanted to do in 1997, that I did get to do, I wanted to understand how technology turned into products and profits in businesses and industries. Because I'd been in the research and technology push side of it for everything in my career to that point. And so I was very specifically looking to go into a company where I could do it from a corporate setting and learn about that. And I absolutely got to do that for the next 15 years.
And so how did the work at Raychem come together?
I got hired as their senior vice president and chief technology officer. That meant I had a modest center research budget and lab and number of people - I don't even remember the numbers now because it's a long time ago. And also sort of a coordination role across the R&D that was done within each operating unit. Raychem was a publicly traded company with about $2 billion in revenue every year. And so I learned things that I could not have imagined from outside. In government we always said we need public sector funding of R&D because the companies can only work on next quarter.
Well, it turns out we were wrong, because at Raychem, we never worked on next quarter, we only worked on this quarter. It was staggeringly short timeframes, and therefore, in my view, short-sighted. And you know, you think you understand how extreme the problem is when you're on the other side, but only when you're in the company do you really see what it looks like, and what therefore doesn't get done or how limited your horizons are. So that was interesting. I had this great conversation with a Wall Street analyst and you know, I'm the chief technology officer, I run the central research function, so we're all about how do we grow. And we're selling these things to our current customers. What are the things that those customers might be persuaded to buy that are additive or move beyond that? And I painted this picture of what the possibilities were to this Wall Street analyst, and she actually got sort of a little – I sensed consternation, and basically, I realized that to her, if we grew out of our traditional areas, we wouldn't fit in her category anymore, and she wouldn't know how to think about us. She had a model, and if we were going to break that model, that was not good. And if you want to talk about what stunts growth and innovation, the demands of Wall Street -- now I experienced it in a way that was really, really useful. So it was a lot of lessons. I didn't have fun, I hated it, but I learned a huge amount.
How fully institutionalized were you in the federal system where you got to this private industry setting, and it was like, whoa, this is a whole different way of doing things? Was that a rough transition, or not so much?
Yeah, but here's why. It sucked so bad because at one point I wanted to spend about $20,000 on a project, and I had to go get permission from my CEO. I was making oodles more money than I had ever made as a public servant. But I remember I was sitting in my office fuming, and what I was thinking was, it's been a dozen years since I had to go ask anyone's permission to spend $20,000. Are you kidding me?
My thesis advisor had a quality metric, a figure of merit, that was the number of dollars that he would get in a research award divided by the number of pages that he had to write in the proposal. So then I created my own figure of merit, which was how many dollars I could spend to do useful work divided by my personal salary, and that figure of merit has been immense in my public sector life, and microscopic in my private sector life. So the short story is, I loved and felt that I gave my whole to society when I was working in the public sector, and in the private sector I learned a lot that later helped me do good for society again, but I didn't enjoy it.
Yeah. Now, Interval Research. How did that come about?
I left Raychem. I was pregnant with my second child, which is a fun time to be thinking about a new job. I had met the guy who was running Interval Research because we were both on the Stanford School of Engineering Advisory Council. It sounded sort of interesting, we started talking, and he ended up bringing me in as one of two new vice presidents. It was an organization that had been formed maybe a decade earlier by Paul Allen, the Microsoft founder, and David Liddle, who was the CEO who brought me in. And the focus was how consumers use information technology, and it was a research lab. David had been at PARC [Xerox’s Palo Alto Research Center], and so he had modeled some of it on what he loved about PARC. And it was at the juncture where the organization had been built around the idea that it was not going to be tainted by thinking about markets and products, but it was going to think in a more unfettered way about consumers using information technology, and it was spewing out really cool research papers. But now I think Paul particularly, was my sense, was interested in, "Okay, let's get somewhere with this!" And David brought me in to run Interval Research and then brought in another VP to focus on how do you commercialize? How do you either license or spin out or do projects that get this out into the world. And that was Doug Solomon, whom I loved working with. And I think we were a good pair.
I would tell you in that year, year and half, what I learned is that if you hire a bunch of people and tell them not to think about anything practical, you can't then get those people to go think about something practical. And they and I, we all went off and later did pretty interesting things, but nothing much really was coming, I thought, out of Interval. I have friends who worked there and who still speak with a twinkle in their eye, or with a glow about the research that came out and how often it's been cited. A lot of things that we were playing with are now products that people use and ways of doing things in the online environment. But--
Like what? What's a good example of those products?
Oh, there were a lot of things about how do you manipulate video. Some of them have been used for good and for ill. We were working on, we imagined a future where you might do video calls. I remember people imagined that for 50 years before any of us actually turned out to want to do it, and we had the bandwidth to do it. But we were working on things that now people do a lot, like could you put an icon, a cartoon icon, of yourself up or more simply, could my image be modified so it looked like I had perfect makeup on? We did a cool project that was, could you operate on compressed video and find different scene breaks? And we used Jerry Springer video and the successful demo was that on compressed video of Jerry Springer, you could find every time a fight broke out on stage. So it was actually, it was very beautiful research, but to me it was clear it wasn't going to go anywhere from that. Nowhere significant out of that organization. So that got shut down in 2000. Raychem and Interval were both fairly short stints. Like a year and a half kind of timeframe.
How much do you see these years as you were sort of riding this wave of startups and the dot-com era and all of that stuff? Was that, particularly with Interval, do you feel like you were just right in the middle of all of that excitement?
It was a little peripheral. It was a little research-y. And it did spit out some dot-coms. Raychem sold specialty materials that were based on some really cool cross-link in chemistry. Heat-shrink tubing for the world - $2 billion of incredibly interesting and diverse markets. But it was a chemistry and materials and component supplier, so it never really participated in the Silicon Valley ethos to a great degree. And then Interval was a little bit closer in terms of the areas of focus, but it was more research-oriented, so it wasn't really until I got to venture capital that I felt that I was in the mainstream of Silicon Valley, I would say.
And were you looking for a move at that point? Did they reach out to you, U.S. Venture Partners?
Interval shut down in...when? The spring of 2000, probably?
And I took some time. And what I was thinking about doing was starting maybe a seed stage venture activity to try to tap into the university research base that I knew from my Washington adventures. And in the course of that conversation, David Liddle from Interval was now, he had left Interval long before the shut down, and he ended up going to U.S. Venture Partners and he said, "Well, why don't you come by and talk to some of my partners?" And in the course of those conversations, that firm said, "Why don't you just come do what you're talking about, but do it inside of our firm?" So I joined the firm in 2001.
What was exciting about this opportunity? What kind of things did you think you were going to be able to do there?
Venture capital is the heart of the Silicon Valley model, and so I was thrilled to get to go be part of the venture firm and understand that business, and really understand how you find great entrepreneurs and how you evaluate business plans. And because this firm was early stage venture capital, a lot of what you do is rolling up your sleeves and working with these really dedicated, creative entrepreneurs to figure out what might be a shot in a business, and then building a business. And so it was an invitation to go right into the belly of the beast, and I thought that would be great, and I did learn all of those things. And again, experiences that you understand, you think you get them in your head, but you don't get them in your soul until you go do them.
Right. So of course, venture capital runs the whole gambit of industry. What was in your portfolio? What areas did you focus on?
Things related to semiconductors. So some of it was semiconductor components for consumer electronics, some of it was semiconductor process technology, and then towards the end, a little bit of clean tech stuff, which was happening at the end of my venture decade.
What kind of stuff in clean tech? What were you involved in?
Briefly involved in a solar photovoltaics company and then the company I really enjoyed was a battery management company. That one emerged because we had an entrepreneur in residence sitting with us at USVP who wanted to brew his next startup. And I started working with him and another VC at another firm, and the three of us became this little, -- before there was a company, we became this little nucleus. And we spent a ton of time together figuring out what we thought the best way would be to do something in battery management. And what we ended up with was something we were all super excited about, and both of our firms seed funded it, and the other VC Abe and I joined the board, and the entrepreneur in residence was the CEO. That was super fun, and that company is still going and has their battery management solution in many, many, many cell phones now, so that's been really cool to see.
Arati, I'm curious, particularly with working with companies doing semiconductors, just to get back to this idea of never letting go and becoming a policy person with a science background, had you stayed on top of the literature at all since your graduate school days? In terms of trends in electrical engineering at the academic and basic research level?
I would say, in the specific areas in which I was working as a DARPA program manager, yes, because we were funding some of that leading-edge work. So it certainly wasn't comprehensive, but in those areas, at that time, I knew rather a lot about what was happening at the leading-edge of research. At NIST, I was in much more of a senior leadership role with a very broad purview.
Going back to what I think I learned from what I studied including physics, the way I validate that a person that I'm talking to, the way my BS detector works, is I ask enough questions to go really deep on some aspect of it. There's something that I'm not usually getting about why they think something can work, and then I really want to go deep in that and then if they can follow me all the way down, and explain it to me, and then come all the way back up and put the whole picture back together again -- that's when I start building confidence in that person. So that's what I did when I was a program manager at DARPA. That's what I did in every management and leadership role I've had.
And so you actually have to understand, aside from the physics and the engineering itself, you actually have to understand the methodology of interrogation of that kind of a technical matter, and you have to understand what is and isn't quality evidence. You have to understand what is and isn't a good hypothesis. The one thing that I liked that I learned from physicists at Caltech is the idea that you examine a concept at its limits. And that's actually something that generalizes beyond physics, so you want to stress-test it. Almost every interesting problem is embedded in a very complex system, and it's almost never the case that you can "just" do one thing, but you examine one particular intervention at the limit. I'm talking about an intervention such as in K-12 education. You can examine a concept and a proposed solution, an intervention, at the limits - like what would it be like if every child in America behaved this way or got this treatment? And in asking that question, you figure out where this idea is good and where it's not good. So that idea of examining things at the limits actually generalized for me in a way that's been very helpful.
So obviously, this is a skillset that you drew on in the venture capital world, looking at semiconductor ideas as well? In other words, you didn't need to be up on the latest scientific literature as if you were a professor working on semiconductors?
Different questions, different questions.
There, what you needed to know was different. We started a company that was doing millimeter wave integrated circuits in conventional CMOS technology, straight-ahead silicon. Beautiful work. So you could bring that to consumer devices. So there, what we needed to know was -- we did need to know who the academic researchers were, working on millimeter wave CMOS. Which was a whole little sub discipline, because we wanted to hire the graduate students and we wanted to find out if they were doing something that was competitive. We might want to harness some of their research, or draw them in some way. So again, an area by area, it became important. We did some work on lithography when I was at USVP, and I had done a lot of, some of the best stuff I did when I was at DARPA in my early years was lithography. So I had been out of it for a number of years, and I remember going to the Stanford bookstore and buying the most recent lithography books to catch back up. So in each area you wanted to know, but it’s very different then if you’re the researcher.
So by the end of your venture at USVP, were you thinking maybe jump back into the world of Washington? Were you thinking about specific opportunities?
No. No I was done, done, done with venture capital, loved what I learned, did not enjoy Monday mornings and partners meetings. I had great partners, but it’s a terrible business model. It’s just a horrible way to function. I know very few happy venture capitalists.
Isn’t it, I mean part of it is that the world–– I mean it’s based on like batting averages are incredibly low. Even the most successful batting averages are quite low.
That’s OK. That’s one of the few things that it shares with DARPA. I love that environment. I love the swing for the fences part. I felt that what I want to do next was something that would make greater impact at a different scale and I wanted to go back a little bit more upstream in the innovation pipeline. I mean, venture capitalist sounds like it’s about technology. But it is about business. You are part of the financial industry and your job is to make money. And so I learned about that. But I have to tell you something. Money is this linear, boring thing, and talk about privilege to get to sit here in Palo Alto and say how boring money is. I love having enough that I can do the things that I think really matter, so you don’t want to get too carried away with this. But money’s really pretty boring and linear. I’m super excited about ideas that are non-linear.
Arati, what you're saying is, you're a scientist, that's what you're saying.
No, I'm an engineer, because I don't want to just know, I want to solve and create. You're not going to talk me out of that, so what can I tell you? So I wanted to do something different. I left USVP at the end of 2010. I almost went to go run a clean energy scale-up center at Berkeley, because-- and I wasn't going to be an academic researcher, but I wanted to use that as a platform to go pull together the market, the finance, the policy, and the academic research pieces that were about scaling clean tech solutions.
So you were really into clean energy as a research problem?
Yeah, well, I would say as an innovation problem. I was not interested as much in how do you improve the efficiency of the solar cell junction. I was really interested in how do you scale it?. So hit pause on that idea, though, because I almost went and did that, but then got the call to go back to DARPA, at which point Pat looked at me said, "We're about to pack up our teenage daughters and move back to Washington again, aren't we?" And I said, "Yeah." I mean, thank God I married someone who understands public service, and the fact that you don't say no when that call comes.
So that's what happened, but you know, everything I was thinking about and wanting to accomplish at Berkeley is now the climate part of what we're trying to do at Actuate.
Yeah, I was going to say, I mean I wonder--
It was fun to pick it up again.
Jumping into the middle of the Obama administration, I wonder if you sort of intuited right off the bat that there would be energy innovation opportunities at a place like DARPA, where you might even be able to make a bigger impact?
Not DARPA. ARPA-E had started, and was underway. Arun Majumdar was the first, founding director of ARPA-E starting in, must have been 2009, at the beginning of the Obama administration. And I was on his brain trust group. I wanted them to succeed. DARPA was about $3 billion a year at that time. Now there was another place in the Department of Energy, ARAP-E, that was spending $300 million a year, which is about the size of one office at DARPA, on clean energy. So DARPA no longer should be in the business. I wanted them to succeed, and DARPA had plenty on its plate. So we did almost nothing on clean energy. I may have had some legacy projects, but we didn't really, that wasn't a focus at DARPA. But then I picked it all up again when we got back to Actuate more recently.
So you got the call, you went to DARPA. Did you feel like you were coming to a brand-new place, or did you sort of pick up where you left off?
Absolutely everything had changed. The Cold War had ended, we had been in ground wars in two parts of the world for a very long time, something you couldn't imagine in my first tour at DARPA. When I left DARPA, not everyone had an email address, and it was just becoming fashionable to put your email address on a business card. Boy, had the world-- the technologies had changed, geopolitics had changed. The people at DARPA, largely, had changed. Of the civil servants at DARPA, half are long-tenured, administrative. They make everything work, so that the other half, the program managers, the technical cadre, they come in for short, 3-5 year years, and the reason they get anything done is these folks over here that are making the machine run. But if you'd been gone 19 years, as I had been, that means that the technical cadre at DARPA had turned over many, many, many times.
So everything had changed, and yet within the first week, I realized that the culture and the spirit was exactly the same. People were still running to work in the morning. A guy stopped me in a lunch line at a salad place across the street my first week. Tapped me on the shoulder, introduced himself. He said, "I'm one of your program managers." And in the lunch line, he started pitching me on his program because he wanted more money. And I was like, "I'm home again! I recognize these people." And the work that was going on was fantastic, and we had a lot of work to do, and I had a great team to go do it, and a mission we all really believed in. So that was without question the most satisfying and I think in time, it will prove to be the most impactful slice of my professional life.
And the question of the extent to which things changed and the extent to which they stayed the same, what about DARPA's place within the DOD generally? Had that overall relationship more or less stayed the same?
Yeah. I will tell you, that and many internal business processes were actually stronger when I came back to DARPA than they were during my first tour. Which is remarkable. I don't know any organization, public or private, where things work better over time. So that's a very deliberate... It takes a lot of work for things to get better. Entropy drives everything to work worse.
So that was really remarkable. At the time that I came to DARPA, it was, I mean, you mentioned that the NIST director role is a Senate-confirmed position, now at the Under Secretary level. The DARPA director position is not a Senate-confirmed position. It reports to an assist-- At that time, to an assistant secretary that reports to an under that reports to a deputy that reports to the Secretary, who reports to the President. And then there are two houses of Congress, appropriations and authorization, and Democrats and Republicans.
And I would tell you that 100% of those bosses and overseers and appropriators understood DARPA's mission, and even more importantly, they understood that we needed to chart our own course, that you don't have a place that's responsible for creating technological surprise by tasking it. That you have to give us room and autonomy to operate. And if you have that and a good mission and great people, that's why it was such a productive and fantastic time. So those relationships were very, very strong and helpful when they needed to help and they got out of the way when they needed to get out of the way. It was great.
So as opposed to NIST, where you're appointed by the President directly and presumably where you would have had a much more direct relationship with the Executive branch, as head of DARPA, you're not getting that sort of direct connection to the White House? It's all going through the Pentagon?
I wouldn't interpret that as reflecting a degree of connection to the White House. The DARPA position is a political appointment, it's not a Presidential appointment with Senate confirmation, but it is a political appointment. So in that process, both the Defense Department and the White House had to agree in order to bring me in. And so I definitely went through interviews on both sides. And then the key relationships in the White House, for NIST they were OSTP and the National Economic Council. For DARPA, they are OSTP and the National Security Council, sometimes. And occasionally, there's a counter-terrorism thing that happens because of ISIS, and then we plug into that because of the work that we're doing to counter online terrorist recruiting.
So there are other points of contact, but generally, I would say in both cases, it was primarily OSTP. In both cases, I got to meet the President, which was fantastic. In both cases, it was clear to me that the President under whom I served actually knew what my organization did. That's just one of the best moments in the world. I remember every word of the microscopic conversation I had with President Clinton about NIST, and with President Obama about DARPA. Those are some of the most treasured memories. But you know, the most fun part was going back and telling my organization about that. Can you imagine? The President actually knows what we're doing and cares and is supporting us. Let's just go, right? It's fantastic. But we were buried. Both those organizations are buried enough. Which is, I think is really good. It lets you do your mission, it's not politicized, and still super constructive relationships.
And so Arati, when you got back to DARPA and you sort of got the lay of the land, how much of your mission, would you say, was about inheriting a really well-run organization, and just continuing a trajectory, and how much was it, there were opportunities for you to identify areas to improve?
Every leader there always has the job of improving the core processes and the relationships with the Pentagon-- there's steady work that you have to be doing all the time, that's more in the continuous improvement mode, right? And then the other half is, you know, if DARPA were working on the same things in 2012 when I returned as it was working on in 1993 when I left, then it would just be a disaster, right? The whole point of DARPA is to run programs in certain areas for a short period of time. To go far enough to either make something happen and have it launch, or make something happen and then realize you need to do the next program that further advances it, or find out that you can't make something happen, and stop and let's focus on something else. And so there's a constant shift and change in the portfolio and in the mix of program managers.
So typically in a year at DARPA, we would start about 30 new programs, but that means about 30 were ending every year. Those numbers-- No, I'm not sure I have my numbers right on that. The full portfolio is supposed to be 200 or 250 programs, so it's probably per year, there were more programs. But it's about 100 program managers, and about 25 or 30 coming in every year and the same number leaving every year. So yeah, you're trying to make the engine work better and better, and you're trying-- my deputy and I spent a wonderful percentage of our time saying, "Look at $3 billion of what we're doing, are we making the best possible portfolio of bets relative to our mission?" That's what we were thinking about all the time.
And so this question of new programs being emblematic of how DARPA has changed with the times, and also how it has been a driver of change over the times. When you came back to DARPA in 2012, what were some of those programs that were especially indicative of the fact that DARPA really was at the vanguard of so many of these changes?
At the time I returned, just about the time that I returned, our neuroscience programs had progressed to the point that a lot of work had done about understanding signaling from the motor cortex and building prosthetic arms with much more sophistication than the simple hook that had been the standard of care for a long time. A lot of this work was started from the concerns of wounded veterans coming back with lost limbs, as you can imagine. But it went very deep on the neuroscience of motor control. And about the time I came back, we were just at that point starting to do our first clinical trials with one of our very early volunteers. She was a woman named Jan who was paralyzed from the neck down, and she volunteered to have these small chips implanted on her motor cortex from which electrical signals could be picked up, and in real time interpreted, so that she could think and control a prosthetic arm. Which was, it was the kind of thing that when you saw it happen in the video, your first reaction was, "That can't be." Like where are the wires, or did they-- are you kidding me? It turns out we can do it, and that was mind-boggling. And just one example of the things-- there are many other things happening in biology. At the moment, it's hard to imagine that there's anything more important from that time at DARPA than the work on infectious disease.
Did you happen to see the Washington Post article? I think just last week? It basically talked about the work that DARPA has done in radically changing vaccines and antibody treatments, and how that led to the first vaccine candidates that went into trials for COVID-19. [Paul Sonne, “How a Secretive Pentagon Agency Seeded the ground for a Rapid Coronavirus Cure,” Washington Post, 30 July 2020.]
Was pandemic research on your radar at DARPA?
Absolutely. When I walked in the door, Dan Wattendorf, who was an Air Force colonel working in infectious disease, had started some programs in that area, and he said, "1918's going to happen again, we don't know when. We're screwed if it happens. And this is how long it takes to develop a vaccine." He was harvesting what was changing in the biological sciences and in biological technology to accelerate. And then he made some incredibly imaginative leaps of his own, like the idea that instead of growing a big antibody molecule to give to someone as a protective measure, let's figure out what the nucleic acid code for that is, and then you don't have to have a big bioreactor and chicken eggs to grow these big biomolecules. You can now give that protective treatment in the form of a nucleic acid string to the individual, and then our own cells build the biomolecule. Crazy thinking which is now actually the kind of stuff that's happening.
So he had started, and we massively expanded, those programs and I mean, that story in the Post actually documented a lot of that pretty well. But more generally, neuroscience, infectious disease, synthetic biology. One of the things we did while I was there was put all those pieces together into a new Biological Technologies Office. I would tell the story about how we started the Microelectronics Technology Office, out of the research office at DARPA back in my early tour. Like I told you, it kept growing and finally we said, "Look, this is the 1,000 flowers blooming office, but now we have a redwood tree. Let's go make it its own office." And we basically did that with biology 20 years later. So that was one whole story.
There's a very massive, and I think ultimately, very important story about changing the way we build and adapt military systems and military system architectures. That was really about trying to not just think about the technical architectures that would be very different, but it was about trying to completely change the sclerotic processes and methodologies for building new military systems in the Defense Department. A lot of really good work is blossoming from that right now. And then in the information technology arena there was a lot going on about cyber security and also driving and then harvesting what's happening with machine learning and artificial intelligence. So lots of things going on, but those were a few of the highlights.
Arati, I'm curious. It seems to me that if you're talking about at DARPA working on improving military information and technology systems, that seems like that's an obvious thing that would be in your purview. But when you're talking about programs that have an obvious biological component to them, how do you figure out or how do you ensure that you're not funding programs that are redundant or are not better-served at a place like in NIH, for example? How do you go about making those decisions?
Well, let's start with why we would be doing it to begin with. And then the question is, well, is there actually a gap and someone's not doing something, or a useful extra contribution we can make. So the reason to do it, you know, DARPA started in the immediate aftermath of Sputnik. That was a surprise we did not enjoy experiencing, and DARPA was a recognition that, yes, we had R&D for Army, Navy, Air Force, Marine Corps, but those were all people who came to work to develop the things we knew we needed. We need a place where people come to work to come up with the things we don't even know we need. That completely change the game and are surprises. And in that context, a part of what DARPA does is try to solve the same military problems that the Defense Department knows about, but just solve them in a radically different way.
The other thing that we think about is what is going to change national security outcomes that this machinery in the Pentagon isn't even thinking about yet? And you know, the role of biology is largely in that category. There are some places in DOD that understood why biology mattered. Biological attack is always a concern. And interestingly, I remember talking to people like General Dempsey who was Chairman of the Joint Chiefs. Army general. He understood that when his soldiers had the flu and were contagious, that it was a readiness issue. So there were places that actually got all of that. But by and large, the Pentagon thinks about fighters and aircraft carriers. So you need DARPA to be thinking about some of these other areas, and in our Defense Sciences Office at DARPA, I always thought of that organization, this is that research part, its job was to look at the research environment. Everything happening across all areas of research. And what they were looking for were places where things are really bubbling, and it's very, very exciting, and it's not Nobel Prize exciting. It's not just basic research exciting. But in the bubbling pot of research in that area, you could start to imagine something that could have incredibly interesting or important national security implications.
So imagine J. C. R. Licklider in the early 60s thinking about computing, and imagining what became the internet. At that time, no one understood what that had to do with national security, but he could tell that something was going to happen that was going to be so disruptive and that's why DARPA ended up doing ARPAnet and the internet. So one of those areas in our era, I think, is biology, because of its convergence with information technology and with the physical sciences and engineering. And so for sure, the bubbling pot is happening. For sure you can imagine important national security implications.
Now, there is an NIH that spends $35 or $40 billion a year, almost all on basic biomedical research. They call it many things, but that's pretty much what it is. And so you might ask, well, what's DARPA going to do? To my great amazement, the community that NIH feeds believes that it is scorned and underfed despite the fact that its budgets have gone up continuously. The reason is that they have a birth control problem, and they keep creating more and more graduate students at a pace that can't keep up with even the quadrupling of NIH over a couple decades. So they all looked at DARPA and said, "Great, you're going to do biology. You'll fix our funding gap." And I said, "No, we're going to do 1% of what NIH does," so for darn sure we're not going to do one more percent of what they already do. God bless them, but I don't think that's actually going to solve these problems that we care about. You would be shocked to find how hostile NIH was to things like our infectious disease work and our neuroscience work. They are in the business of publishing papers. We were in the business of demonstrating solutions that could change the world, and that is all the difference in the world.
And there's another six hours to say about what solutions R&D is and how it's different than basic research, but you need both. You absolutely need both, but biology and biomedicine is an area where we have gone way long on basic research and we aren't thinking in terms of solutions in our problems. In my era, it was only about $300 million a year in biological technologies at DARPA. But you know, they were sitting on top of this big basic research base, and it was doing something that no one else is doing.
Especially by the second term of the Obama administration, I remember myself just, this is the time in which five star [generals] would start talking about climate change as a national security issue. Which they hadn't really done before. It was a pretty significant thing. I'm curious from your perspective, what did DARPA have to offer in this broader idea that climate change is, it's not just a political issue, it's not just a human rights issue, it's also a national security issue? What role did you see that DARPA had to play in these bigger questions?
This is a case where it's almost the inverse of what I just said with NIH, because again, because ARPA-E had a clean slate, $300 million a year, and we wanted them to succeed. They had a budget that was already at a level far more than we would carve out within DARPA. Specifically on that issue. So again it really just wasn't a particular focus at DARPA.
Did you want it to be? Was that something that was sort of like out of your control?
No, I mean I had-- I could do whatever I needed to do at DARPA. No one ever told me not to do anything or told me to do anything. I had tremendous latitude. I just didn't see what we could add to this great stuff that was going on at ARPA-E. I mean, they were friends, we talked with them. There were occasionally, you know, there are a couple of joint projects, but if we couldn't be additive, what was the point? It just didn't make any sense.
Arati, the wonderful little institutional history you provided about the origins of DARPA and Sputnik and things like that, from the outside looking in, it seems like there's an inherent contradiction of DARPA, that it needs to be nimble and it needs to be ahead of the curve and it needs to think outside the box in all of those things. And yet, obviously, it's smack in the middle of the world's largest bureaucracy.
(laughs) You got it.
That's exactly the term I use to describe DOD.
So you know, it's only at your level, you know, directing DARPA, do you fully appreciate that inherent contradiction, and so it would seem that day in and day out, you have to solve or beat that contradiction in order to do what you need to do.
I didn't have to. And it's because all of my bosses and all of our overseers and appropriators on the Hill, they got it. To me, it's an existence proof that even in the world's biggest bureaucracy, you can create space. You need space and you need support. You need both. You need to be an island that is separate from the morass of the bureaucracy so that you can create something new, and then you have to be able to build a bridge back to it if you're going to actually implement the new thing that you came up with. So this organization, the parent bureaucracy has to give you both space and support. And if the Defense Department can do that for 60 years, then anyone can do it. Like absolutely anyone can do it.
Can you give an example of a particular initiative where in identifying this constellation of support that you have, on the Hill, in the Pentagon, what would be an example of, here's a program that you want to see to completion. And who are the people that you need in order to see that through, who are allowing you to do it in as nimble and innovative a way as you want it to be done?
Hmm, okay. Yeah, let me give you the arc of a particular program at DARPA. It might be the best example, or an example, to get at that. A little before I arrived at DARPA, there was a program manager who had been a Navy officer. He had a great idea. You can have a remote-controlled little boat already. Why don't we build a ship, a full ship, that it's not remote controlled. It's highly autonomous. It's basically a self-driving ship, with very minimal supervisory control. It can leave the pier and navigate across open oceans for months at a time. He thought it was a great idea, DARPA thought it was a great idea. The Navy did not think it was a great idea. In fact, they thought it was such a terrible idea that they tried to get DARPA to not do it. My predecessor wisely said, "Thanks for your input," and went right ahead and did it.
By the time I got there, the Navy had gone from hostile to merely skeptical. One of the things we did at DARPA was every-- it was called a quarterly, it only happened maybe two or three times a year. But we had these regular standing meetings with the vice chief of each service. So this is the number two military leader in each of the military services. So when we had our meetings, our standard meetings with the vice chief of Naval operations, the first one I had when I got to DARPA with that individual, they had gone from hostile about our self-driving ship to merely highly skeptical. And that's a really important stage, because it's in skepticism that they tell you all the reasons they think it won't work. So now you understand what you really have to do to prove it out. And we kept going, and we started proving out the things that they were skeptical about, and before I left DARPA, I got to christen the first ever self-driving ship that came out of our program. At the time that we did that, we were in full partnership with the Navy, and we've now handed that ship off to the Navy, to the Office of Naval Research. Now they're going to take this thing and figure out, how do you best use it? Do you trail after quiet Russian submarines, in a cheaper way than we can today? Do you use it for dangerous missions like clearing sea mines? These are the things that we had in mind, but now you've got a prototype and the Navy can start testing it, and seeing operationally, what would they actually do with it?
And I haven't followed it closely, but I think they're moving forward and buying more systems, which is good because it'd be great if they bought more of those. What I really want to see is that they start changing the concepts of operations because they can have this much more inexpensive asset that's very independent, doesn't require a lot of staffing to control, and doesn't have any human beings on board so it's, as we say in the DOD, it's “attritable”. If you lose it, the world doesn't end, right?
So that's the journey. But if you had to get the permission of the department or the Navy to start, you'd never get there. And on the other hand, if they never paid attention or if you never got them past hostility to engagement, then we would just build a cool thing and nothing would happen. Does that help? Answer that question?
Absolutely. That's a very vivid-- I mean you would never get that story except the way that you told it, so that's really terrific. Arati, were there some things looking back at your tenure directing DARPA, that you really wanted to change, either culturally or scientifically? That you just were not able to do for whatever reason? Is there anything that sticks out in your memory?
I'm looking at the clock, and I'm wondering. I've got a one o'clock and I'm going to need a little bit of time before that, so let me hit pause for just a minute. Maybe, I think we should probably do one more question, so I'm wondering if you want to do that, or is there-- I don't know, I think we covered a lot of territory. What do you want to do to wrap up here?
I think actually, I mean, it would be that question, and then the only other thing that I wanted to talk about, because we already talked about your current work, is... Was your time at Stanford sort of an incubator for Actuate? So let's just make the mini question the, anything that you felt like you tried to accomplish but couldn't at DARPA? And then the intellectual incubation at Stanford as a precursor to Actuate, would be those two questions.
Okay. Let me tell you a thread that runs from DARPA through CASBS at Stanford to where we are now, and I think it'll maybe get to all of this. So I talked about in the 80s and 90s the redwood tree that started growing among the thousand flowers, with microelectronics.
In the 2010s, it was biology. The new thing that's still a flower at DARPA but I think is going to be a redwood tree eventually is about social science. An interesting question is, how does DARPA evolve and change direction? And the answer to that is, it's through who you hire. So I decided I wanted to have a biology office and I put some people in charge of that, but when they decided what direction they wanted to go within biology, they would hire program managers in that direction. That's about as much top-down, and then the rest has really got to be bottoms-up through the program managers and the program concepts that they come up with and then they go execute. But it's in selecting those new program managers that we set strategic direction.
One of my office directors-- a couple of them, actually, one in particular, head of the Defense Science Office, started talking about how the social sciences are changing. Partly because of all the data that we have, many other reasons. And as we were talking about, your first reaction is, "What does that have to do with DARPA?" But quickly, if you peel back the layers, it's hard to think of anything that matters more in national security than understanding and trying to influence human and societal behavior. That is what the whole business is.
And so that's true, and now we're seeing a bubbling pot of research where things are happening. It looks like really interesting research, and it might have huge implications for national security. We have got to be there. So we hired, while I was at DARPA, we hired our first PhD anthropologist as a program manager. And he started some really interesting early programs. But what that did for me personally was it just got me thinking about this change happening in the social sciences, and I was super intrigued by it. I thought it might have something to do with these societal problems that I was worried we weren't innovating for, as I was leaving DARPA. Like maybe the social sciences has something to do with how people have opportunities in life, and how health care is delivered, and how we trust information. And possibly even climate.
And so I was really intrigued, but you know, now you know my entire professional life, you know that I had lived among the hard sciences and largely the physical sciences. I'd gotten a big exposure to the information science and engineering area. But social sciences? I didn't know about them. We were moving home to Palo Alto, and I met a woman named Margaret Levi who runs CASBS. In the course of our first conversation, she said, "Oh, why don't you come spend a year at CASBS?" CASBS has been around about 60 years. Every year they bring in about three dozen fellows and during that year, everyone does their own thing and they all eat lunch together and it's a little community tucked up on a hill, just on the other side of Stanford campus. Almost all of the fellows are social science academics who are on sabbatical for a year, but they are very diverse in terms of - some are economics, some are psychology, some are sociology, some are anthropology -- everything you could imagine in the social sciences. And then Margaret likes to stir in a few oddballs, and I got to be one of the oddballs. I told all my fellow fellows that I was doing anthropology on them - I was trying to understand social science and social scientists. And Adam Russell, the anthropologist we had hired at DARPA, came into DARPA, and he was a little freaked that he was the only social scientist surrounded by techies. So now I was having the inverse experience. And we had a lot of fun laughing about that.
And what I-- I learned a lot, and what I learned is, number one, God bless physics, but these social scientists are asking far more interesting questions, because they ask the questions about human beings. It doesn't get any more interesting than that. Number two, they are working on complexity that makes what we do look like child's play. Your subject changes while you are interrogating it. It makes quantum mechanics look like a walk in the park, right? So that's crazy. And as a consequence, I will have to say, I don't think they've gotten that far. And I think that they're not that optimistic about their changes of doing-- I don't think that community feels that it has agency in the world. The way that scientists and engineers, in all the fields I've worked in before, people have agency. They know that what they're working on can actually make an impact in the world, and I don't-- social science has not really got that culture or feel.
So I actually left feeling... It reinforced my sense that there's a huge opportunity that's coming. It underscored that it hasn't happened. And I think making this convergence of social science and the informational evolution, data and data analytics which is, that's just a fountain of correlations. Then the ability to model really complex systems and reason about causality, and then third, the ability to reach billions of people through a device that they carry around with them all the time. So you can now personalize and interact and connect in ways that you could never have imagined before. These are the kinds of ways that I think social science is going to turn and do something that you can grab and use to solve problems. I think it's going to go from Know and Understand to Solve and Create. That's what I'm excited about. As you can tell.
And it's not like we're going to have a model that explains all of human behavior, ever. I'm not talking about that. What I'm saying is that the boundary between tractable and intractable problems - that's the human enterprise, right? It's how do you move that boundary between what's tractable and what's intractable? It's going to accelerate a little bit and that acceleration is going to change millions of lives if we can do it right. That's what I'm really lit up about. And so that became a formative part of thinking through what do you actually do about these big, hard societal problems? And in a sense, it converged with my concern about those problems, and that was then the basis of forming Actuate.
Well, Arati, I'm looking at the clock, and I want to let you go. This has been awesome, talking with you today.
I really enjoyed the walk down memory lane. How nice of you to listen to all of that.
This is going to be, just so you know, we talked at the beginning, you know, who are the kinds of people that will find value in these interviews, and I mean, there are so many different access points to gain value from all of your insights over the course of your career, and it's just Exhibit A why I take a big tent approach to who to talk to in the bigger world of physics, because it is a big world out there, and your engineering mindset, it's so clear how this informs all of the things that you've succeeded at over the course of your career. So, really, I'm so happy that we connected. And thank you so much for doing this.
Oh, it's so nice you reached out.