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.
Courtesy: Marvin Weinstein
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 Marvin Weinstein by David Zierler on April 11, 2021,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
For multiple citations, "AIP" is the preferred abbreviation for the location.
In this interview, David Zierler, Oral Historian for AIP, interviews Marvin Weinstein, Chief Science Officer of Quantum Insights, and emeritus physicist of SLAC. Weinstein describes the origins of Quantum Insights in partnership with David Horn and the development of a data mining algorithm called Dynamic Quantum Clustering (DQC). He recounts his upbringing in Brooklyn, his early interests in physics, and his undergraduate education at Columbia. He describes the big issues in physics at the time, including the two-neutrino experiment, and he explains his decision to remain at Columbia for his PhD to study under Gerald Feinberg. Weinstein explains how he became a postdoctoral student at the Institute of Advanced Study with the endorsement of T.D. Lee to work with Roger Dashen on K13 lepton decays. He describes his subsequent faculty appointments at Yeshiva University and then NYU, and he discusses the opportunities that led to him joining the theory group at SLAC. Weinstein describes his work on PCAC and the Higgs mechanism, and he explains how DQC originated from his interests in quantum mechanics. He explains his subsequent work in lattice field theory and then core and condensed matter physics, and he describes the changing budgetary environment at SLAC over the course of his career. At the end of the interview, Weinstein conveys optimism that his focus on the health industry will demonstrate that the adoption of DQC and its ability to analyze data will lead to better health outcomes across a spectrum of ailments.
This is David Zierler, oral historian for the American Institute of Physics. It is April 11th, 2021. I'm so happy to be here with Dr. Marvin Weinstein. Marvin, it's great to see you. Thank you for joining me.
It's great to see you, David.
To start, would you please tell me your current and most recent title and institutional affiliation?
My current and most recent title. So, my current and most recent title is I'm the Chief Science Officer of a company called Quantum Insights, and I'm also the founder of the company.
Now, you're also an emeritus physicist from SLAC.
And an emeritus to LASLAC National Accelerator Laboratory. I was there for, I don't know, forty-two years. I got there in 1972, which was a super-exciting time to be at SLAC.
And I was there for about forty-two years when I retired in order to found Quantum Insights.
What year did you retire?
Oh, you're asking me a hard question (laughter). I retired probably in 2014. Sound right? That's about forty-two years, right?
Yeah. And in what ways, if at all, have you remained connected to SLAC?
Oh, I'm emeritus. Except for COVID, I have an office at SLAC, I go in periodically to see friends, to talk to people, to keep up with things at some level, keep my finger in one thing or another. But mostly I'm really working at Quantum Insights now.
Marvin, we'll develop this further in the conversation, but was the founding of Quantum Insights related to your decision to retire from SLAC?
Absolutely. About four or five years before that, I don't remember exactly, something like 2009, 2010, David Horn, a friend of mine in Israel, and myself developed a data mining algorithm called Dynamic Quantum Clustering, published on that, and started using it for all kinds of problems. It works in a wide range of areas, and in those days, we started with problems in physics. We started with problems in condensed matter physics and moved out to other things to test it. A new algorithm doesn't mean it beats the world, so you have to work on it somewhat to make sure it really works, and you need examples. That was taking me very far afield from what the current Department of Energy considers to be my job. Now, that's a little bit funny. My job was unusual because I got tenure at the lab in '76, something like that, and I was very lucky. My letter of tenure was about a 10-liner. It basically said, we'd love to have you join the theory group, and your job is to do anything you think is good science as long as you're willing to serve on some committees and things, that's fine, and as long as you chose to publish on whatever you want, that's fine. And so, in effect, nobody could make me work on something I didn't want to work on, but I must confess, my interest became very much what I was doing with DQC because- it sort of sounds a little funny, but it had the potential in the areas of medicine and research in medical areas to save lives, and that was sort of something I never managed to do.
And it was very exciting to me, and so it was really where I wanted to put my time. And it was very clear the lab wasn't interested in that, so if I'm doing my job, I have to do the things I wasn't in the mood to do at that point in time. And so, I retired and went out and formed Quantum Insights, and we've been in business now for about eight years.
And Quantum Insights, as you envisioned, was a full-time endeavor. It wasn't a retirement hobby.
This was going to be your next full-time adventure.
Yeah, yeah. I don't have retirement in me. But my whole life has been going to play every day. I mean, yes, by definition I was working, but in point of fact, my life has never been something where I could say, well, I could shut this off because it's something I'll do even if I don't have to do it. And so, this was my next passion, and so, yes, I just got the opportunity to make it full time.
Marvin, a question we're all dealing with, for you, from both the business side and the science side, how has the past year-plus in the pandemic been with remote work?
For us it works fine. Our company at the moment is three people, and that's all we need to do what we do, and as business increases, probably, we're probably gonna have to bring on some more people just to cover the business. But we gave up our offices at the beginning of the COVID crisis because we weren't going in, and it really hasn't put a crimp in us at all, but we're lucky that way. We can do everything remotely.
Is there an office to go back to once the COVID crisis is over?
Well, there will be at some point an office to go back to probably. It's still fun to meet face to face.
And it does make certain things easier, but really working remotely for us has been very successful.
Marvin, let's take it all the way back to the beginning. Let's start with your parents. Tell me about them.
My parents. Well, I come from a Jewish family in Brooklyn.
Where in Brooklyn?
Where in Brooklyn? Crown Heights, President Street.
So, on one side of the block I lived on was Lincoln Terrace Park. Across the park was my elementary school, PS 189. Boy, you're really taking me back into the past (laughter). And my Brooklyn accent is coming back strong and stronger as we talk.
Marvin, how Lubavitch was Crown Heights when you were a kid?
It wasn't. At that time, for the Lubavitch you had to go up the Eastern Parkway and go past Franklin Ave in order to see the Lubavitch. They certainly were there, but the neighborhood has now completely changed.
Although, peculiarly enough, since I'm able to look on Google Street View, the apartment house I grew up in is still there and it looks remarkably the same. It takes me back when I look at it just because it's the same. Utica Avenue is not the same. It has changed a lot, so the canonical kosher delicatessen, appetizing store, bakery, and Chinese restaurant are no longer there (laughter).
Yeah. How many generations back does your family go in New York?
Oh, no, no. My family were immigrants, so my father was born in this country. My mother came to Ellis Island at some unknown age because she came with her two sisters and a brother, and they had quite a story because basically their mother had passed away, father decided to come to the U.S., and the day they were supposed to get on the boat to come here he died of some fever.
And they put them on the boat, and they arrived at Ellis Island and were met by an aunt who had some job getting them through. And my mother never knew her age because- I can tell the secret now, nobody can deport her, she lied about her age to get in (laughter). And so, when I would say, "How old are you, Ma?" she would say, "I don't know."
What were your parents' professions?
My mother was a housewife, and my father was- well, as anybody who had gone through the depression, he really was a schoolteacher. He founded or played a big role in founding the program for retarded children, CRMD, Children with Retarded Mental Development, in the City of New York. He taught at something that was or became John Jay High School. And then, he always kept his job at Western Union in the evening because you never gave up a job if you grew up in the depression. So, he worked day and night. And he was pretty neat (laughter).
Marvin, how Jewishly connected was your family growing up? Did you go to Shul and Shabbat, that kind of thing?
So, we were- I guess they would be called some kind of Orthodox in the sense that they were, but they weren't all that observant.
They were traditional.
Traditional, right. And the thing was, my grandmother lived in the neighborhood a few blocks away on Carroll Street and she never actually really spoke English, she spoke Yiddish, which I learned in self-defense. It was very important to know what people were saying about you.
And so that was a necessity. And we went to a synagogue. I mean, you're asking me twenty questions. We went to a synagogue which was a couple of blocks from her house, which was called a shtiebel. It was really in somebody's basement.
And maybe the rabbi was a rabbi. I mean, it was never totally clear. He arrived in the country and announced he was a rabbi. And as everybody said in those days, “but he lost his papers on the boat.” And so, he was there. My grandmother loved him, and so we went there. So, I never actually belonged to a temple. We weren't in that class. Which meant that when I had to go to Hebrew school, which I did for many years. My parents, through some strange way, found out about something called the Mendel Folkshul, which they never realized was sort of a communist operation (laughter). They were founded by the Farband Central, and so for years I thought everybody in Israel referred to one another as "comrade." Everybody was Chaver Schwartz or Chaver Gubkin. And so, it was an interesting experience. And I went to Hebrew school religiously until I was- not that I cared about it, but it was a great place to go. And then, after classes were out, a friend and I would go around the corner into a vacant lot and pick up two sticks and be d'Artagnan and fight away. That we didn't kill each other was remarkable.
I would never let my kids do what I did. And that went on 'til- I spent even a couple of years in Hebrew high school on weekends and one evening. And then I said, I don't have time for this. And that was fine. Nobody cared.
When did you get interested in science?
I probably decided- I tell this story and my wife cracks up, I actually decided I wanted to be a physicist I think around- or scientist, around age eight.
Not that I knew what a scientist was. I was at that time already fixing the family TV when it broke. That was no great thing. I had the courage to take the back off the TV, having watched the repairman come and fix it. And I noticed that inside the TV was a piece of paper that showed you where all the tubes were, and it had labels for what they did. And so it wasn't really hard when a tube burned out to look at the TV, look at the label that said picture or sound or whatever, take that tube out, go around the corner, have somebody test it, buy a new tube, and put it in. So, I was the neighborhood TV fixer. It was not any depth of knowledge. And basically, I usually say around age eight. I had been running around the house with a towel around my neck jumping off things and trying to be Superman. Age eight I was probably the age when I figured out it wasn't gonna work, and I wasn't gonna grow up to be Superman. And then I just wanted the next best thing, and as far as I was concerned, the next best thing was being a scientist.
Where did you go to high school?
You did go to Brooklyn Tech?
Tech was spectacular. I have great love for Brooklyn Tech. It was a very interesting school. And probably at Brooklyn Tech, the biggest influence on me was my English teacher, Mary P. Heslin, who was fantastic. And the second biggest was my physics teacher.
Who was your physics teacher?
I'm trying to remember his name. I'm really bad with names. He had a PhD, and he was the math teacher. Forgot his name right now. Bernie, Bernie, Bernie Sachs. It was Bernie Sachs. And I think one of the nicest things was at some point when I was at the Institute for Advanced Study as a postdoc, we went into New York, several of us, to an APS meeting at, I guess, the Hilton, who had no love for physicists at meetings like this because physicists didn't spend money on the appropriate things. They went into the bar and they sat there with one beer for five hours and wrote on all their napkins, so it was not good. And at any rate, he was giving a talk to the American Association of Physics Teachers. And so one of the most fun things, we went over, because a friend of mine also came from Brooklyn Tech, and so we went over to see his talk, to see what it was, and we let him know who we were at that point. He remembered us and he was so excited that we had actually become physicists. And I was at the institute, and my friend was out at SLAC, so he was tickled pink. It was really a great afternoon! (laughter)
Were you thinking specifically physics programs when you applied to college?
Yeah. Yep, absolutely. And [Sachs] was the one who pointed out to me when I was applying to schools and I was thinking of various places, he said, "You know, Columbia has a really world-class physics department." And of course, they did. They had Rabi, they had T.D. Lee, they had Jack Steinberger. There were all kinds of people who were and would become famous and important. Leon Lederman was there at the time. So, in the end, I went to Columbia.
Did you commute from home or did you live up there?
I commuted from home. We didn't have the money for me to live at school. So literally it was crazy. I'm in an ivy league school and, basically, I was spending $700 a year for classes, or $700 a semester. My New York State scholarship covered that pretty much. And so really it wasn't hard. And in my sophomore year I got a job. Previously, I had worked in the Catskills during the summer. Standard New York story. And then, that year, I saw an ad on one of the bulletin boards and they were looking for people who were interested in working for some company called Basic Systems, and it looked interesting. So, I answered the ad, and it was a lot of fun. That was the first teaching machine programming place that had ever existed. Three psychologists from Schering Drugs, I think, had been teaching pigeons how to do things. And then they decided, well, pigeons and outside salesman are about equally easy to teach. So, we'll start writing these programs to teach these guys to sell things and bring them back every several weeks. They would take a several-hour teaching machine program. And then teaching machine programs meant paper, right? There were no computer programs at that point doing any teaching. And the idea was to write these things down for teaching. So, my job was to write something to teach high school students plain geometry.
Marvin, what were some of the exciting issues in physics that you remember as an undergraduate. What was going on in the department?
Well, at the time, remember, the two-neutrino experiment had just been done. The standard model was starting to peek its head over the horizon. People did know what was sort of going on, although QCD didn't exist at the time. Nobody knew about the gauge theory link. Murray Gell-Mann had done SU(3), and so that was exciting at the time. And so that was all being discussed. Some of the most fun, of course, at Columbia was seeing Rabi. And Rabi was not exciting 'cause he was either a good teacher, which he wasn't. He was awful. I mean, Rabi really barely got through a lecture, and then the next day, 'cause he didn't get through the lecture, he'd give it again. But he was just a fun guy. He was an interesting man. To me, he looked like a little Jewish tailor who I always expected to be running down the hallway with a tape measure around his neck and a piece of chalk in his mouth. But he wasn't. He actually was an extremely sharp guy and politically connected, involved in tremendous things too. And he had some issues with his prostate, so at the time you actually spoke to Rabi, there was this fiction that he's invited people into his office for cider and donuts. That was never something that happened to me. But we spent many good conversations in the men's room while he was trying to take care of his business (laughter). And he would capture somebody to have a conversation, so you would get caught if you happened to be there.
Marvin, did you know an Ephraim Fischbach as an undergraduate?
I did. We were buddies. Yeah.
Did you know him from Brooklyn also or just from Columbia?
No. No. Ephraim I met at Columbia.
What were some of your favorite classes as an undergraduate?
Favorite classes. Actually, I was a little weird. So, I took a lot of math. It's not weird for theorists now, but at that time I took my first math course from Samuel Eilenberg, so he was at Columbia.
So, he's the father of abstract nonsense, as he liked to call it. Categories and functor theory and things like that. And I remember in that class several of my compatriots that I got to know well, we had all placed out of first year math and we didn't want to, so we decided what we would do is we would take the math major course in math rather than the physics major course in math, and then that would be ok. And so, we signed up for his class and came in and his class was over-subscribed, and he came in and he said in his inimitable way, "This class is too large. It'll be a lot smaller in a few days." And then he said, "I understand there are many physicists in the room." He says, 'I want you both to know if, by the end of this course, you know a vector has anything to do with physics, I will consider myself a failure" (laughter). He didn't scare us away. We actually all did well in the course, but it was quite entertaining!
Now, Marvin, you just liked the math, or you did this because you thought it was good for theoretical physics?
I liked it. I just liked it. No. I rarely did what was good for me. I followed what gave me pleasure, and in this case, I found it very interesting. And so, I took a lot of math through differentiable manifolds and things, and that was good for me as a physicist, because in my course on differential topology, I had- I forgot his name. He was Scottish, was the teacher- Dold. And he said, "There's this conference going on in Santa Barbara." And I said, "Well, I can't afford to go to a conference." He said, "No, no, no. We can get you money to go to the conference. And it's on relativity, topology, and groups." And I was very interested in general relativity at the time. And so, I got sent out to this meeting. And in those days, meetings were a lot nicer than meetings are now. I really lucked out in many ways. I grew up in a period where things were smaller. Nowadays in big meetings people show up, they give their talk, and they disappear. If they're the people you want to meet, if you have any interest in meeting them, you have no chance if you're a student. But at that meeting, everybody with their families was living in the Santa Barbara dorms, and it went on for two weeks, and you were together for meals and you were together to play ping-pong, and you were even together to play poker with some of them. So, I ended up as an undergraduate playing poker with Wheeler and Misner and Ray Sachs and Alfred Schild, all of whom were terrible poker players. So, that even I could make money on the five and ten-cent game we were playing, and I did. And it was amazing! But they were really exciting to be around, really interesting. And so, it was a great experience.
Marvin, did you think about leaving Columbia for graduate school, or was there a particular professor who said stay?
I did. It's really funny. I, in fact, applied to three places. I got into all of them. One was Princeton. And in the end, I turned Princeton down. They all came with fellowships and so it wasn't a money thing. For a weird reason I liked relativity. I had gotten to know Wheeler, and I had the feeling that if I went to Princeton I would work on relativity. And I had decided at that time already I was more interested in particle physics.
And so, I said, well, really good particle physics group here, really good particle physics group there, but I think I'm gonna go get caught in the relativity group down there. And so, I stayed at Columbia, which was great.
Was there a GR group at Columbia?
Not really, no. So really, it was great.
What were the other two schools you applied to?
I think I applied to MIT and maybe Cornell. I never thought about the West Coast at that time.
Mm-hm. Who ended up being your advisor as a graduate student?
Gary Feinberg, Gerald Feinberg.
That's what I started to say, but he went by Gary. And he was T.D. Lee's student. So, I worked for him for, I guess- I was in and out in a little less than four years.
Did you know him as an undergraduate or really only as a graduate student?
I knew him a little as an undergraduate. Mel Schwartz and then Gary were my undergraduate advisors. So that wasn't doing too badly.
So that worked well. In graduate school, I also got friendly with T.D. Lee, and I think it was his recommendation that got me into the Institute for Advanced Study.
What was T.D. Lee like?
Oh, I loved T.D. What was T.D. Lee like? Very smart, number one (laughter). Extremely smart.
But approachable, you could talk to him?
Oh, yeah. Well, I don't know for everybody- I tend not to be afraid of people so, yes, I could talk to him. In fact, sometimes you could even make fun of him. He was a very nice man. Many really good physicists have this characteristic: He would tolerate a lot. He wouldn't tolerate bullshit. Am I allowed to say that? take it out, right?
So T.D. and Feynman, Feynman was even more like this, if you're giving a seminar or something, he's good to have in the audience. Feynman was a great guy to have in the audience. He would ask a question, and as long as you answered it honestly, and "I don't know" was a perfectly good answer, but if you tried to spin some story that is made up out of whole cloth to cover up the fact that "I don't know" is really your answer, then, as I would say to my wife, there was blood on the floor.
At that point, you got ripped to shreds. So, were they nice guys? They were until you crossed certain boundaries. Those boundaries had a lot to do with intellectual honesty. So, if you did that, you paid for it. It wasn't fun.
What was Feinberg working on by the time you became his student?
Well, actually, it was interesting. I went to work for Gary- remember, my story was I didn't want to work on relativity.
But the year I started actually working after my first year in graduate school with him on a thesis project, he had gone away for the year that I was still doing- and met Dennis Sciama and became interested in relativity. So, he had me do several problems with him in general relativity, which ticked me off. And then I did problems in particle physics. I did those, I showed they didn't work, it wasn't going to make anything for him. And then I did what I to this day believe was a terrible thesis. It was a ridiculous problem, but you did it. And so, I hid my thesis. I don't know. I tried to get all the copies from the Library of Congress, but I don't know if I succeeded (laughter). And then went off to the Institute for Advanced Study where I started to work with Roger Dashen on what was called PCAC.
Marvin, you won't get off that easy with me. What was your thesis on?
Oh, you don't want to know. It was a crazy idea. There was an incorrect experiment by Pipkin that saw a resonance in electron-photon scattering. It really was obvious to us that it probably was nothing, it was a mistake. But Gary said, "Well, why don't we try to do-" and this is something that's used for dispersion theory. There is a topic called dispersion relations and it was used in strong interaction theory. Geoff Chew, and people on the West Coast were doing a lot with it, Mandelstam. And my advisor said, "Well, let's try doing an N over D (N/D)-" and I'm not going to go into an explanation of that calculation of electron-photon scattering "and see if there's a resonance." And I said, "But it doesn't really apply because the photon is massless, and the particles have to be massive.” And he said, "We'll work our way around that," and whatever. Did the calculation. It was fine. It was a lot of calculating. It was very annoying because it was tedious. It was the sort of thing T.D. could do in his sleep. T.D. was the most organized calculator I've ever seen. He would take out a very large sheet of paper and he would start at the top, and without erasing he would work through the most-lengthy calculation and finish at the bottom. And that was it. It was stunning to watch. That was not me. Never has been me. And I sort of said to Gary, I said, "Look, this is really very annoying and, anyway, I have to have numerical values for these functions,” which showed up in massless particle scattering called Spence functions. And I said, "They don't exist in any books. They have to be computed numerically, so I should learn to program." And he said, "No, no, no. We'll figure out how to do analytical estimates." Now, it was clearly not a great idea, but what happened was he again went away for six months, and the day he left I got myself a book on Fortran and got myself an account with the computer department, a computer account- I forget which machine it was. I learned all the joys, you really learned in those days to be tough, because you were always faced with the Hobbesian choice, right? You had a very big deck of cards that you carted off to the computer center. And now, the first thing that confronted you was, did you run it through the duplicator? 'Cause there was a fifty percent chance the duplicator would eat your deck, then you would be punching it back in; or did you not, and give it to the computer guys who would drop it on the floor, or have it eat your deck; what did you do? And so that was your everyday thing. And so, I made many, many trips until I finished my thesis. My poor girlfriend at the time had to make her one a.m. stop at the computer center for me to check if the program ran or not, and I would get lost for forty-five minutes or something if something had happened so it could be resubmitted so it would be ready in the morning. Those were the good old days. When I got to SLAC and found out you no longer had to do things like that, it was a revelation. It was unbelievable (laughter). At any rate, yeah, so I finished my thesis that way.
Marvin, besides Feinberg, who else was on your committee?
Oh, my committee. My committee was Feinberg, T.D. Lee, and Sam Devons, an experimentalist.
Was it common to have an experimentalist?
Always at Columbia, there's always an experimentalist on theory.
Columbia worshipped experiment, and so, yes. Actually, the funny thing is Devons was the one- I remember it was really funny, 'cause he looked at me, and Devons was a Brit, and he had this ability to appear to be clueless about something. Of course, he wasn't. And he asked me in the nicest- I answered T.D.'s questions, I answered Gary's questions, and then Devons said, "So, Marvin, tell me, why does the neutrino have zero mass?" I looked at him, I said, "Well, are you asking me what are the best experimental limits on the mass of the neutrino?" And he said, "No. Why does the neutrino have zero mass?" And I said, "Are you asking me why it's possible to write a theory where renormalization doesn't give the neutrino a mass?" "No. I want to know why the neutrino doesn't have a mass." Now, T.D. Lee is on the floor. He is laughing. He's hysterical. And at that point, I couldn't have told you my name. I mean, I had gotten to the point where T.D.'s laughing at me in the front row of this little office, Devons is asking this question. I haven't got a clue what's on his mind. And finally, Devons, with a very innocent look, looks at T.D. and says, "Do you think this is an unfair question?" And T.D., laughing, says, "Well, I've seen very good people make idiots of themselves on this subject. Let's see what Marvin has to say." At which point Marvin could not say anything.
I mean, if you'd asked me my name, it was not gonna come out because I wasn't comfortable. But then Gary said, "You did fine." So, they put you out of the room for ten minutes and you hear laughter, waiting with the secretary, and then Gary opens the door and says, "You passed." He said, "Come, you need a drink. Let me take you out to the faculty club" (laughter). So, he took me to the faculty club. And I said, "Gary, I have two questions. T.D. asked me this question that- I answered it all right, but I really didn't know why he was asking me the question. And Devons was a joke." I said, "I don't know why the neutrino has zero mass. For all I know, it could have a mass." I was ahead of my time.
There's no reason why it has zero mass. And he told me what was on T.D.'s mind. "Well, T.D. just wanted to see if you'd read this paper he wrote fifteen years ago, and that was what his question was about" (laughter). And Devons' just was Devons. Sam likes to ask questions like that. So that was it.
Was the opportunity at the Institute already wrapped up even before you defended?
Yeah, I think so. You're asking me about- you know how old I am. You're asking me questions about- I mean, the dinosaurs had just died at that point.
(Laughter) But you don't have any specific memory of applying for a lot of postdocs?
I remember, and I believe it occurred before the exam, so it was only gonna be a question of if I'd be allowed to go. But T.D. came down the hallway and said, "You're gonna go to the Institute for Advanced Study." In fact, he said, "You are Oppenheimer's last act."
And it became I killed Oppenheimer because, literally, Oppenheimer was very sick, and there was a faculty meeting. And this is in Dyson's book, in fact. So, basically, he said, "Take Weinstein, he's good." And then he walked in his front door and dropped dead, which is why I say I killed Oppenheimer.
And so, I never got to meet him. So, I knew in advance that probably as long as I passed my exam, I'd be going to the institute.
But that was the plan, that you would be working with Oppenheimer?
No. No. I'd just be going to the Institute.
In fact, the person I worked with who was invaluable was Roger Dashen, who unfortunately died young. But Roger got me involved on something that became the beginning of my career. For about three or four years that's what I worked on, which is partially conserved axial current (PCAC). The realization that there was such a thing in particle physics, (and that goes right to that time) as a spontaneously broken symmetry. I always prefer hidden symmetry or Goldstone symmetry because it became terribly difficult to keep talking about a really broken, spontaneously broken symmetry. It sounded stupid, but that's what it got called. And so, Roger was great, and the next three or four years of my life that's what I focused on. That's how I became a phenomenologist.
What was the overall intellectual environment at the Institute like?
It was very good. It was really good, especially among the other postdocs. We had a great time together. And seminars were very good. They weren't what they were in the- I mean, Regge was there, and Dyson was there, and Rosen, but he was really in plasma physics. And Roger Dashen and Steve Adler. So, things were exciting. I grew up in physics at a tremendously exciting time, right?
That's just before the standard model is coming into existence. And all hell is breaking loose. And so, it was very exciting. And Dyson was just fun to get to know.
Would you spend a lot of time at the Princeton physics department?
Most of the time was spent at the Institute. We would go to the Princeton Physics Department for seminars, and they would come up to the institute for seminars, so you got to know people that way. But, no, I didn't spend a lot of time at the physics department.
And was Dashen's work, was this entirely new to you before you got to the institute?
Oh, yeah. I didn't know anything about it, it was great.
And what was exciting about it? What was happening with Dashen at this point?
Well, he and Steve Adler had both become permanent members of the Institute at a very young age, at just that year. So, he was super smart. Roger was one of the smartest people I've ever known. And he was just fun to talk to and fun to work with. It was very much a collaboration. It wasn't he was my boss, or he told me what to do or things like that. No, we had knockdown drag-out discussions about various things, and just a great intellectual environment. And the work was really seminal in certain ways. We were the first ones at that point to strongly make the case that what had gone under the name of the Goldberger-Treiman Relation, which at that time, thought to be just a consequence of the low mass of the pion (from dispersion arguments.) was really due to something else. And it was derived on the basis of an assumption which wasn't true, and instead we made the argument in that first paper that it is a real consequence of the fact there was an underlying symmetry that said the pion mass had to be zero if the symmetry was unbroken, and that the symmetry was actually broken. We didn't know about quark masses or whatever, but it was slightly broken, and that was enough to give you the Goldberger-Treiman relation. So we laid out the technology for starting to do all of these calculations and showed how it could be derived from a phenomenological Lagrangian, how you could invent a phenomenological Lagrangian based upon low-energy theorems that, in fact, would be exact if the symmetry was unbroken.
What experiments may have been relevant to this line of research?
Well, for us the most exciting one was the one I learned about the day I walked in the front door of the physics department at SLAC, which is a friend of mine from graduate school, Dave Hitlin, a very successful physicist, experimentalist, who's now at Caltech, and he said, "Oh, I gotta tell you something." So, Roger and I had proven a theorem on the form factor in Kl3 decays, that is, the K lepton, the Kl3 lepton decays. And we made a prediction about that, and he said, "Guess what? You're dead on. It's right. But there's something a little funny about the experiment." And so, I looked at the experiment and he says, "We're having a little trouble here." And I said, "No, you're not. You're having a little trouble here because you're violating Lorentz invariance, in this result right here." So we had a talk and I looked at things and I said, "I think it's the way you're doing your radiative corrections, 'cause you're making the assumption that you have a four-Pi detector and you don't have a four-Pi detector, and that makes the calculation wrong." So, he actually did the calculation over and, in fact, it fixed everything, and everything made perfect sense. So that was, for me, the most fun, 'cause that was a real prediction about a real physics problem made years in advanced being correct, and that was fun. There's nothing quite like that feeling, that you've really seen something about the real world, and you got it right. And it's all nice to have abstract theories, I mean, a kind of- sorry, these string theorists, they have all these lovely, wonderful theories, but no predications about the real world that are correct (laughter). So, it's fun. That was great fun. And I continued pursuing that. I went to Yeshiva for a year and then two years at NYU, and then I went out to the West Coast.
Now, the two postdocs, essentially the three postdocs, to what extent was this a function of the job market at the time?
The job market was horrible at that time.
After I finished at the institute, there were very few jobs. And so, going to Yeshiva really was- Roger basically told me, "Well, don't worry about it. Write two letters and you'll have your pick of the job, and just decide where you want to go." And I wrote two letters, and I didn't have a job. And Lenny Susskind, actually, was the one person at Yeshiva that hired me, so I spent a year at Yeshiva.
And it was a pretty good department.
It was a great department at that time. There was Lenny, there was Yakir Aharonov, there was Al Cameron, Jim Truren. There were very good people. It was a little weird. Yeshiva was a very strange place in the sense that, and they eventually got a very fancy new building, but at that time the theory group was in a catering hall, what was an old catering hall. And there was the mezzanine of the catering hall and then there was one floor further up. And there was a very lovely old woman who really- she was the cleaning lady, and she was too old to do the job. So, whenever you walked in, you were in dust bunnies up to your knees. It was very funny. But the place was exciting, and lots was going on at that time, so it was fine.
And NYU was a good experience also?
Yeah. So I segued at one point into formal field theory, and that was because Wolfhart Zimmermann was at NYU and he had just worked all this stuff out that I really liked about a formal structure of renormalization, which was easier for me to understand the technical details of, how to prove things. And I was still worried about PCAC, but really from the point of view I wanted to understand how the photon got a mass or how any particle mass- at this time already t' Hooft's stuff had come out and I wanted to understand all of this stuff. And so, I started working on that problem, and I was- I had learned Zimmermann's stuff and I decided, well, I was gonna do this particular sample problem for myself as an exercise. And I was on my way to SLAC, and we were driving cross country, my wife and I, and were stopping at Fermilab. So, I was at Fermilab for six weeks or something like that, and then I continued out to SLAC. And while I was at Fermilab, I worked on this thing, finished it, and it turned out, in talking to Zimmermann- well, he and John Lowenstein had been working on the problem, too, and it was really kind of funny because my proof of how it worked was very much a physicist nuts and bolts thing, and theirs was correct in the sense they dotted all the i's and crossed all the t's. And I had a simple physics argument for why you could get a certain result. They had a formal proof. And they said, "You know what we should do? We should just write a paper jointly because we have the same result." And so, I got to write a paper- I actually got to write three papers with Wolfhart Zimmermann. I wrote the first one, which presented things my way, which most people could understand, and then, the next two papers, they wrote, which did it all correctly with all the epsilons and deltas and everything else that needed to be there for a rigorous proof. And so, I had a wonderful time working with them, so I enjoyed my time there. And I had one other thing that I'm lucky for, which is a lot of fun. I got to know Engelbert Schücking, who was a general relativist, and we talked. And we had an actually very funny experience. I guess it's ok to tell you this, and you'll decide what you want to say about it, but you'll be amused. So that was, in fact, what got me to SLAC. I took my NSF grant money and went to a physics conference at Cornell in the summer of the year, and it was the outgrowth of this- I had done it just before I went up there, I was very bothered by Joe Weber's experiment. I really wanted to know- he had this gravity wave detector and I wanted to know: How do you calibrate a gravity wave detector? How do you know why it's responding? How big does the signal have to be? And Weber had a theory paper on that, and I think it was forty-five pages. And to me it felt kind of hard to buy into everything that went into that paper. And I realized before I went to the Cornell meeting that, in fact, there was a really three-line derivation of it. And so, I wrote a little paper on this very short derivation and sent it to Phys. Rev. or Phys. Rev. Letters, I don't remember who it was, and it got rejected. And it was very clear from the rejection who had written the rejection because the rejection said, "This is an extremely clever paper with an amusing result that shouldn't be published." Pretty clear. And one of the reasons was I didn't refer to the unpublished thesis of Joe Weber's student. Now, all right, I got that rejection while I was standing at the mailbox. Now, I got the rejection, and I wasn't feeling good about it, so I was standing at the mailbox and talking to Engelbert Schücking, and I said, "Look, Engelbert, I'm a particle physicist, I'm not a relativist. I don't do this for a living. It's not important for me to have this paper published." So Schücking said, "It's a good paper. It should be published." And I reach into my mailbox and there's a letter from Freeman Dyson. I had become very friendly with Freeman, and I would send him anything I enjoyed, so I sent him this paper, and I said, "You might enjoy this paper." And the letter that I got back from Freeman said- and I still have it somewhere- it said, "Dear Marvin, I really enjoyed your lovely paper. It's so much better than what I've written on this subject. Will you please check the following calculation your way?" And Schücking said, "We got this solved." He immediately goes to the Xerox machine and xeroxes the Dyson letter, staples it to the rejection, and we type a little note that goes on top that says, "I would accept the referee's opinion but for the feedback we've had on this." And the paper was published the following week. So, if you have the right friends in this world (laughter).
So that was one of the more fun things that happened to me in my existence.
Marvin, what were your initial impressions when you got to SLAC?
Oh, center of the universe.
No. In fact, I went to SLAC, really, I'd been at NYU for a couple of years, and I was looking around for jobs. The market still wasn't great, and the jobs that were available weren't all that exciting. And I had met B.J., Jim Bjorken at this conference. I'd met him, I'd met Feynman, and I'd met Ken Wilson at that thing, and spent a lot of time talking to them. And I called B.J. up and I said, "Look," I said, "I should be looking for a job, but I'd love to come out to SLAC. I've never been out to the West Coast. Very exciting things are going on at SLAC. It's just I will not come out for a two-year postdoc. The appointment, whatever it is, has to be for a minimum of three years because I don't want to arrive and immediately start looking for a job. I want a year to sit quietly and think about the problems that interest me." And this is the difference in the times at that time, B.J. said, "Give me five minutes." And he hung up, and in five minutes he called me back and he says, "It's done. This is your salary. This is what you'll do. And, yes, it's three years with no problem."
So, visiting scientist was really a glorified fourth postdoc, essentially?
Really, yeah. I mean, yeah. You knew it was a visiting scientist. Good for you (laughter). Yes. It had never been done before, and it was amazing. It just happened in fifteen minutes. And then I never had to write a job application again, so I was a happy soul.
What group did you join?
The theory group. Sid Drell was head of the theory group. There only was one theory group.
How big was it at that time?
Oh, SLAC was small. That was when SLAC was really at its height. I mean, you just had scaling, B.J.'s work; quarks were new; everybody wanted to be visiting at SLAC. So, in the summer when people were off, I was telling my wife, it was like the most exciting place because everybody, Sidney Coleman, you name it, would come out.
And literally it was a very ecumenical kind of thing. Sidney would arrive and not have a desk for two weeks while he was waiting for somebody else to leave. And in the meantime, there was a long table in what we called the green room, which was the seminar room, where all the guests would sit and have a little place on the table that they could write. And, of course, the corridors were full of people talking to one another, yelling at one another, writing on the blackboards. Everybody was confused about the physics that was going on. It was just heaven. SLAC was a dream come true. It was just fabulous.
Did you ever have opportunity to interact with Pief?
Oh, yeah, absolutely. Pief was fun. Sid Drell was wonderful. These were all terrific people. They were great. It really was terrific. I interacted with Pief less than, for example, I interacted with Burt Richter, who followed Pief as lab director. Burt and I interacted a lot.
What work did you do in the theory group initially?
When I got to theory group, I followed up on the- I'm just trying to remember really. I wrote more on PCAC and the Higgs mechanism at that point, the work I had done with Zimmermann and company. I got very interested in lattice gauge theory, and in particular, nonperturbative methods, 'cause clearly the issues that I had cared about were nonperturbative. So after a couple of years, Sid Drell and Shimon Yankielowicz and I started doing things on what got called- Jesus, I don't remember- it got called at some level the real-space renormalization group, but it was a method for approximately calculating nonperturbative things in lattice systems in general, lattice field theory. And over time that grew into something that I developed with a postdoc, Colin Morningstar at the time, called the Contractor Renormalization Group, which extended work I had done with Jim and David Horn and Sid Drell and Helen Quinn on these approximation techniques. And that's how I eventually started working in condensed matter problems, because this whole—it looks like I had this career that suffered from a short attention span because I kept changing the formal field I was in. I went from phenomenology to- but there really is a through line. I was following one set of problems. The reason for going into condensed matter rather than focusing on QCD is you're putting forward a computational method which has never been used and saying it works. Well, there's only one QCD, and if you get the right numbers, you fudged it. But there's a lot of condensed matter physics, and it all has exactly the same structure. So, it's full of real problems in which you can make real statements and then are you right or are you wrong? And so, I ended up doing a lot in that area for a period of time, and with some detours from time to time. But that was the focus for a long time. And started working on problems with Assa Auerbach and others in Israel. And then down the road, and we come to DQC, dynamic quantum clustering. And really that's because as I went further and further into this, I started work in nonperturbative methods in quantum mechanics and things like that. And basically five, eight years of that work was what made it possible for me to sit down and write DQC in a matter of a week, the code, because that was going to be the heart of the method we used to make it work. So, it really is a through line, but it doesn't look like that (laughter).
Marvin, what was it like for you to be there during the November Revolution?
Oh, that was fabulous! The funniest thing, of course, is I missed out on a bit of that because my daughter was born on 11/11/74. And so, when the announcement was made of the discovery of the J/psi, I was in the hospital with my wife. And it was tough night, so by the time I got back and fell asleep, I got a phone call from B.J. saying the new particle has been discovered, and I thought he was making a joke (laughter). So, I've never forgotten that day, or my daughter's birthday. But then life was super fun.
What were the circumstances when finally, you got a full-time tenured job at SLAC?
Well, like everything else in my career, it felt weird. I didn't expect to be discussing all this stuff with you. What were the circumstances? Well, it was clear there was a job that was coming up. It was a staff job, but at SLAC the staff position was for theorists; for an experimentalist it was another story- but for a theorist it didn't make much difference. And so, I applied for that job. I was very sure I wasn't gonna get it. And that was what I decided. I decided there was another candidate, and that candidate was definitely gonna get the job. And told my wife that, and that was the way it was. And we had plans to go down to a meeting in La Jolla. There was always a La Jolla a meeting at that time, and plans were to go down there and go to that meeting. That was an electron photon meeting. And the day before I was told I had the job. So, yes, very excited. But even funnier was going down there and after the meeting friends of ours, Myron Bander and company, had a brunch at their house before we started to drive back to SLAC. And Ben Lee was at that dinner. I knew Ben. And basically, Ben said, well- I hadn't been bothered because not only was I sure I wasn't getting the job, but I wasn't drowning in job offers. They weren't pouring in. And Ben said to me, "So how are you feeling?" I said, "Well, grateful, 'cause I really didn't have very many offers." And he said, "Well, we knew you were going to go to SLAC, and if you got the offer, you weren't gonna leave." So, yeah, they didn't bother (laughter). I said, "Thank you very much." I didn't know I was gonna get the job. Could've relieved my stress level a little bit. So that was my wonderful story. It was fun. I was very excited. It was great. I was offered other positions which a good friend at one point said, "Would you like to come here? We have an opening." And I said, "Why?" "Well, you'd be a professor." And I said, "I'm really happy where I am." So, I stayed. That was that.
What did you do next, given that you had this security and sort of freedom to work on whatever you wanted?
That's exactly what I did next. I worked on whatever I wanted.
Which was what? What new projects did you take on?
At that point, I was pursuing lattice field theory and pushing that really hard. And then I was pushing core and condensed matter physics. But that was already starting to be something that would later with the DOE become a problem. I was hired to do particle physics and as far as they were concerned, I was doing condensed matter physics. I had my letter of tenure that said I could do whatever the hell I want. In the later years, towards the end before I retired, that was becoming difficult with the DOE. In fact, I was getting pressure to not do what I was interested in doing, which was, at that point, DQC. Persis Drell and I had a conversation. She said, "Well, the client doesn't want this." And the lab was not gonna support it. And, yes, I was fine with my letter as long as I stayed in particle physics, which was not what my letter said, basically. But there was real pressure at that point even to go back to writing papers on a subject I no longer thought was worth writing papers about or do whatever I could do to get out. And my attitude was, I said, "I'll tell you what I'll do. I want two years of not being bothered by you to finish up what I want to do and getting this out the door, and then I'm gonna retire." And so that was the deal. I was left alone completely for two years. Nobody bothered me. Nobody told me what to do. Nobody pressured me to write papers on the approved subjects by the DOE, and then, yes, I was more than ready to retire at that point and start Quantum Insights, so I did.
But, Marvin, wasn't the larger story at this point that SLAC itself institutionally was moving beyond particle physics? Weren't your changing interests sort of illustrative of larger changes?
But it wasn't going in the direction the DOE was funding.
You have to realize, when I got to SLAC, I think Burt told me there were four lines of funding. SLAC now has 100 or 200 lines or something. The evolution from physics as I knew it at SLAC- you could give me an example of something that happened in the old days at SLAC. At a certain point, I guess it was in 1980- now I'd have to go look back at things, God knows- I was really annoyed with typing physics papers and even on the computer, printing them out, pasting in pictures. It was terrible. I became aware of something at the time called TeX. Donald Knuth had written this computer program which made it possible to do very good typesetting on a computer and then print it out on a laser printer. I got a copy of the program. I had it sent up to the SLAC Computer Division from Stanford. And we couldn't send it to the laser printers 'cause the software didn't exist for doing that, but I could send the output to Versatec plotters in the administration and engineering building, so, that proved that we could to it, and the output was great (laughter). And at that time there were some people on the computer coordinating committee, which was run by Louise Addis, who was fabulous. And so, I went into that committee and I pushed for this to become an accepted thing at SLAC, to be supported. I said, "This will be great. It's going to change the face of the way we put out papers. Nobody's doing it yet, but it will happen. This is the way to do it." And Louise was very open to it. Les Cottrell, who was more conservative, said, "We don't need it. We have this IBM script." And I said, "It's crap. It doesn't produce publication-quality output, but TeX does." And, "Oh, but there's this macro package for it written by Gary Feldman." I said, "I'll make a deal with you." At that point, Keith Rich and the man who wrote Top Drawer, Roger Chaffee- I said, "We can't get it to the laser printers." And they said, "Sure we can. We can do it in a week." And so, I said, "All right. By the time it's ready to go to the printer I will have a macro package for you which will do letterheads, and which will do papers, and which will do this, and will be easy to use for physicists." And I wrote something called PHYZZX, P-H-Y-Z-Z-X. And if you look in the literature, PHYZZX is the first thing that exists there. Later, there was a competing package, which I never cared about a competitor. That was wonderful. And then LaTeX came out, and my attitude was, at that point, I had supported PHYZZX in the community for five years, and it was a lot of work on the side. So, I was really happy when LaTeX came out. People said, "Do you miss it?" and I said, "Absolutely not." everybody is using TeX now, that's the way it's done. So, yeah, I played a significant role in bringing TeX to the physics community. It was the first place it was getting heavy use and things like that. And you could do that at SLAC with no effort. You just had to go to the computer committee and make your case, and then the world changed. Now, it would be 400 committees and three years of effort and money to get the same thing done that was done in two weeks.
Marvin, to establish sort of an intellectual history for the origins of Quantum Insights, how much of it was related to your career as a physicist and how much of it was, these were sort of extracurricular activities, interests that you were pursuing beyond your world at SLAC?
So, I love the way you're asking that question. It's not so easy to answer because your life is never- at least my life is never that easily separable. The person who came to me after all with the idea for part of the solution was David Horn who invented QC, quantum clustering. We take the data and use it to make a topographical map of the data. And then we have to take the data points and use that topographical map to associate them with one another in the appropriate way. He was a longtime collaborator on problems and real-space renormalization group, but he left particle physics before I did. Started working on various things. One was neural nets. And he would come every summer and we would talk. And he showed up with this idea and we did our usual thing, we screamed and yelled at each other for hours on the blackboard for about three days, arguing about what it meant. And at the end it was, oh, that's what it means. And then the question was, how do you make it possible to do this thing? And then I said, "Well, I gotta do that, 'cause that's exactly what I've been doing in my nonperturbative methods in quantum mechanics." I had all the software, I had all the tools, it's just putting it together for this problem. So literally it was a week. So, is it connected? Yes. Quantum mechanics is not a fundamental part- quantum mechanics has zilch to do with clustering data, but it's a trick which lets you do what you otherwise wouldn't be able to do by gradient descent or other methods like that. And so, yes, it played a huge role, but not in an expected way.
Was there anybody at SLAC that you talked to about these things, or you kept that sort of private?
Other than David, no. No. It wasn't that it was a secret, nobody cared. Nobody cared. I've told you this. It's sort of very funny. When you do something that's really-everybody always wants an idea that's outside of the box. Actually, nobody wants an idea that's outside of the box.
And this kind of thing isn't outside of the box, it's outside of the building you keep the box in. So, it's really not what anybody does, and that's way too unsubstantiated for people to want to spend their time learning about it. And anyway, they think they already have things that work as well, maybe. You have to make that case, okay? That's the uphill climb, and it's worse out in the real world when you're making a product. And when you're dealing with biologists, which is a whole new thing, and the medical community, the drug companies, the standard of proof they have for things working was very tough. It's worse, right, because they could kill people. And so, it's an experience. It's a learning curve for sure. I'm in a place now that is very different from what I grew up in. And physics now is very different from when I grew up. The physics world now is much more straight-jacketed. People don't get the freedom to do what I did. I mean, changing fields constantly, moving around, working on what interested me, you can't do that in physics now. The funding agencies don't let you do it. There was this question, why don't you go to NSF and get funded in this thing? They have these calls all the time. And I laughed. I tried that for a while, and I said, "You want to talk about something really—" You're getting me to dump on you all the things that annoy me—but I just thought this was the most ridiculous thing I ever lived through was call for a proposal in data mining. The call for a proposal says we have no way to do this. We don't know anything about this problem and don't know how to solve it. We want proposals on how to do it in the following way ... What? You just said you haven't got a clue. Why are you telling me how to solve it? The only proposal you'll accept is to solve it in a way that doesn't work. What's the matter with you? What part of your brain is not functioning? It's ridiculous, but that's the way a lot of it works now. It wasn't like that.
Intellectually, academically, what was the division of labor between you and David?
Well, David doesn't program so for sure that was- the original idea for the function was his absolutely. At that point, of course I did the programming, of course I did that part which he didn't know how to do, but the truth is, the development was- I can't tell you where one thing leaves off and the other begins. It's always a back and forth and back and forth and back and forth. I don't know. I wouldn't lay claim to anything more than it was a collaboration. It was like that. All my collaborations have been like that. Saying I did this, or they didn't do that, I don't know that. I wouldn't be working with the person if it wasn't shared effort.
What was the learning curve like for establishing a business? You had been in academia this whole time.
Nutso. In fact, it was very good. It didn't do well in the beginning. It was much harder to move these things out into the real world than a naïve person believes. I funded the business. And I'm lucky, I did well in life, and I was retired. I didn't need an income and I was able to invest something in the company and work free. At some point, one of the advisors that I got suggested something called the Alchemist Accelerator program. We joined Alchemist. Well, first I had to pitch Alchemist, which was an experience because, if you look at my early talks and my later talks on DQC, they start getting less high-brow, more low-brow, shorter, not short but shorter, and I was faced with the fact I had a five-minute pitch and that was it to get into Alchemist Accelerator. It worked fine. Got in. And it was a great program. And actually, my son-in-law- Bernard Chen didn't have a CEO, so my son-in-law, I asked him to join the program with me just 'cause you're supposed to have two people in the program, and he did. I think we had a very good time and learned a lot, pitched and raised close to a million dollars in ten days or something to fund the company for a while. A lot of it came from the physics world. Burt Richter was an investor in the company, bunch of people. I shouldn't be telling the people's names. And so basically that worked fine and there then became the problem of actually figuring out how you're going to solve this. Everybody originally thought, well, it will be software. Really not a good play. First place, it doesn't make any money, but second place, it's deceptively easy to use the technology. The technology could be learned by anybody in a short period of time, and you could just do it. To understand what you're doing and what it's telling you, that turns out to be not easy. You have to change the way you think about things. So, just handing it off to somebody without training them for a long time, they'll do all the wrong things. So really, it's telling them not to do certain things and hitting their hands when they go to do it, 'cause that's the thing they do.
When did you get outside confirmation that DQC could be monetized, that there were real customers out there who would be willing to pay for this?
I'm not sure we still have that confirmation, ok? (laughter) Did certain jobs that worked well. What the right way to monetize is not really clear. Our current business model we would like to- well, part of it is I pushed it in that direction, but it's a good place for us, is biomedicine, things like that. Cancer is a narrower focus. All basic diseases that can be studied genetically and things like that are certainly things- autoimmune disease, all of those are good for us to work on. So, then the question is, how do you make money in that space? As a consulting house, yeah, we can make money. It's not tons of money but, even worse, it doesn't penetrate as powerfully as we would like it to, so that it makes the biggest difference in people's lives. And so much cooler would be to develop companion diagnostics and get a piece of companion diagnostics, which will then fund the next round and so forth so that we work collaboratively with drug companies or some research facility and get some ongoing revenue from that and some money for doing the project. And are we getting traction in that area? Maybe. It's slow. We've learned a lot about the bio area, the cancer area that we knew nothing about, but even more about how you have to talk to businesspeople. It ain't the same world as science, that's for sure. It's not academia.
How do you convey to a business executive that this program, this approach is actually good, not only for the bottom line but for better health outcomes?
So, the answer there has been- well, the good news is I don't have to do that 'cause my son-in-law stayed as CEO and it's his job to do that. But the real answer is in the drug area, drug companies, things like that, it's hard. Literally nobody gives a damn that you work on physics data and do a better job than anybody. Nobody cares that you can analyze hyper -spectral camera pictures to see hidden tanks in the desert better than anybody else. They have no interest. They only care about one thing, their data. If it doesn't work on their data, they couldn't be less interested. So of course, they all want practice runs. They all want you to do some low-paying practice. And it took a while to learn really important lessons, one of which is it's ok to do those, you don't have to make a lot of money, but if there's not a lot of money involved in them, you need something else that will get them involved in working with you. But they just don't hand you something they don't care about, right? You have to make sure it's a problem that means something to them, to their bottom line, because if they just hand you something they don't care about, they don't pay any attention. They did it for some curiosity reason, but they didn't do it because it's gonna get you any traction by the time you finish, so in that case you're wasting your time. You're just wasting your time and money doing these things. And so, where we are good right now is, we have a pretty successful collaboration going on with this company, Cepheid where we have, in fact, demonstrated we can see things in their data they can't see. Their technology doesn’t see what we see, and our stuff is actionable and important to them. So, we're now going into some longer-term set of projects which will end up being good for them and good for us. We're in a project with Genentech, which will certainly result in our being a partner on a paper, but it's also demonstrated that we can do things that are actionable and important. So, really I'd say we're at the point now - from a business point of view - it's good 'cause we're in a repeat business stage. We didn't do a demonstration project and can ‘t get a second project, we got the second projects and third projects, and that means, yeah, people are realizing we're for real. Now, once we're able to brag about those things, we'll probably be doing better after that. So, the answer is, it's been a climb. I think we're starting to get real traction at this point. The process has been one of doing demonstration projects and then showing if we can knock it out of the park. And it's good, but you really have to make sure that they're gonna work with you, because if they don't care, their guys aren't gonna be talking to you. We see something but we don't know what that is in terms of their product, in terms of their stuff. For that we can tell them what we saw, but they have to interpret it. We can give them all the information, but they have to know if it's important to them or not important to them, useful to them or not useful to them. And so, we need their buy-in. It has to be they want to collaborate with us, not that they hand us something and say, go tell us what it is. Ok? That doesn't work. And so that's an educational process for both sides, and it's the challenge of doing what we do because we're not doing what the world does. We're doing something extremely different, and we're saying things about data and how you should look at it that is not the party line. It works, works really well, but it's not the party line. And so, it's not what they're really looking for Just like the DOE is not really looking for a solution, they're looking for a solution that fits their parameters for what they'll accept as a solution.
Marvin, I know the business executives don't care, but I certainly do.
What are some of the hallmarks of a physicist's sensibilities that you can see in the DNA of DQC?
You want me to analyze the DNA of DQC?
No. Coming from a physicist, like why did this not come from a computer scientist or an engineer?
Oh. Because to really do what we're doing, you do have to have a deep understanding of quantum mechanics. I'm a weird person. I'm a little oddball. I really know quantum mechanics. As I said, some of this is unpublished work that I did that I never thought was interesting enough for the physics community to write a paper on or didn't meet my standards for a paper, and then I also am pretty good at programming things. So, it's that mix that isn't out there. It's much more likely to come from a physicist because physicists know how to program, but few programmers understand quantum mechanics. It's not gonna happen. It's why I'm not really worried. I've often said, and it's not proven, we could give a programmer the code, they can copy it in hours, they can reimplement it in another language in hours, they better not change a line of code 'cause they won't know what went wrong. So, we're kind of protected in that sense because it's pretty hard for you to go out and find a set of programmers that also understand what's going on in quantum mechanics and why we're doing the things we're doing. So, it had to come from a physicist. Maybe a quantum chemist. It'll be okay that I call them physicists, since they are. It's not really a mistake (laughter).
Do you see DQC strictly in the realm of classical computing or in the future is quantum computing going to really change what DQC can do?
You're gonna make me a bunch of enemies. I am not at the moment a huge fan of quantum computing, ok? Quantum computing to me at the moment is the place where nuclear fusion has been since I've been in high school.
It's gonna be there in ten years. It's been a long time and we still don't have commercial fusion. Why? Because as you push it forward and get to the next level of performance, plasmas do something you didn't expect. Quantum computers will be the same way. Just as somebody who worked in condensed matter physics and given how they work, they're gonna do surprising things as the problems become more and more complex. It's a whole new area, right? As things become complex, phase transitions occur. So, in a system, after all, take a high-temperature superconductor, we know everything about those things. There's no surprising elementary physics which explains high-Tc superconductors. It's Coulomb's Law, the exclusion principle, and some chemistry. Everything fundamental is known. The problem is there's lots of stuff going on. And lots of stuff goes on, that stuff ends up doing things that are very surprising and we don't know how to compute when a phase transition occurs; some new, surprising property emerges, mostly unexpected. So, think about the fact that I can't compute that a desk exists. If I tell you there's a desk and its measured properties, I can compute the hell out of what you need to do to build that desk and make things, but I can't prove there is such a thing as a desk. From first principles, knowing everything that matters, Coulomb's law, how atoms work, quantum mechanics, I can't prove there's a desk. Huge- that's the kind of thing that got me interested in nonperturbative physics. You can't do it from perturbation theory. Perturbation theory starts with a problem you know the answer to, and then changes it a little bit and calculates the corrections to your answers. But when something like a desk forms, that's not perturbation theory. Just ain't. It's not perturbation theory. So, yeah, I had occasion at some meeting- it was kind of funny, actually, there was somebody from- not IBM, maybe Righetti, I can't remember which of the companies that are doing quantum computing- was talking about their stuff and I was really interested. I wanted to hear that because I wanted to know where things were. And I heard his talk, and this was a while ago, so they've come a ways from this, and I don't know, so I'm talking about history. I'm not talking about where I know things are now 'cause I haven't been following as carefully as I should in order to make these statements. But this was the incident. I'm listening to him and he's telling me about it, and he extracted the results. I raised my hand and I said, very politely, "So, I worked on these kinds of problems in condensed matter physics, and you're looking at 9 x 9 lattices of computing elements, and then you're extracting the answer by looking at the spectrum when you make a perturbation.” I said, "That's great, but when it becomes 500 x 500, all of a sudden the system becomes almost rotationally invariant and you will not be able to use those spectra anymore to extract the sum." He said, "Who are you?" Ok. Nobody asked him those questions, but those are the things that a physicist knows these systems are gonna do, not because he's so smart, just 'cause he's seen it before. You know that's what's gonna happen if you've worked on these things theoretically. They get complicated and the whole thing, called emergent phenomena, stuff happens- that's the right answer, stuff happens. Okay? (laughter) That's it. And you don't know what's gonna happen necessarily. And so, what you have to do to fix the stuff that happens it's gonna be the next job. So, somebody says why don't I believe in quantum computing? I say, "It's invented by Feynman. It's gonna work." Okay. I have no doubts. But Feynman never gave me a timescale. Ten years? I don't think so. Forty years, maybe. I don't have the time to wait forty years. I'm just too old for that, so it's somebody else's problem, not mine. So, I'm off from your question 'cause I don't want make a prediction. I don't believe quantum computers anywhere in your future or my future are gonna be general purpose computers. You won't want to land your plane based on the fact that the quantum computer has managed to land it. Quantum computers do certain things provably better than regular computers. It's a limited class. And the important thing is, for example, if you're factoring things into primes, quantum computers can be maybe 0.999 likely to give you the right answer. That's not a guarantee you're gonna get the right answer. You might have to try ten times to get the right answer. The important thing about the quantum computers that might take you ten minutes to try ten times but finding it exactly by conventional means might take you ten years. So, you can afford to be very wrong as long as your life doesn't depend on being right.
Marvin, now that we've worked our way right up to the present, for the last part of our talk, I'd like to ask a broad question that sort of puts it all together. So, you already told me about the through line, even though your research agenda has been very eclectic, right?
Do you see that through line going through DQC, and if so, how might you extrapolate these interests ultimately to determine what success, long-term viability for DQC looks like?
Well, the through line is always where my nose is pointing. I mean, I'm following my nose about what interests me. Sure, long-term, well, I told you what my measure of success is. In cancer, I want to make some contributions which play a big role in saving people's lives. In many ways that can happen. For my partners' point of view, they also wanted to make money. Yeah, I like money. I will never turn down a million dollars if somebody offers it to me. I have no principles about that. But I certainly- that's not my motivation. And since I have very nice partners and they tolerate me, as long as they can be interested and making money, they're happy, so that's great. And I think we all want to do good; and doing good and doing well are even better. Was it Khrushchev who said, "Communism is great, it's even better with bacon fat.”? So, yeah, it would be nice to make money at it, too. So that's, of course, a validating thing, somebody's willing to pay for it. But more exciting is we'll actually save some lives. I can imagine that when they let us get out of the box a little bit, when the constraints of business free us to work on a wider range of problems, we'll be versed in diseases other than cancer. I mean, we could make contributions in a lot of areas where- after RNA expression data, proteomics data, understanding what it's saying, how it relates, they're all the same thing for us. And so, for sure there are many areas in medicine where we could play and be helpful. We can't afford to, at the moment, spread ourselves that thin, but in the future if we're successful, sure, I'd love to play in those areas. An area I think is crucially important is autoimmune diseases, and I believe we could make progress there at some point. That's a belief. It's not proof (laughter). I just believe it, okay? And so, yeah, that's where I see it going. It's a wonderful playground from my point of view as a tool, the tool is fun, understanding the tool and its limitations is fun, but being able to use the tool on a wide array of interesting problems, that's heaven. I've always been a bit of dilettante and this lets me be a dilettante in areas where I have little knowledge to begin with, and then, as we've seen results, you learn what you need to know to understand the results. So, yeah, I told you, I'm boring (laughter).
(Laughter) But you might save some lives at the end of this.
That would be the whole deal. That would be the whole deal.
Marvin, it's been a great pleasure spending this time with you. Thank you so much for doing this.
Oh, it's fun talking to you.