An Interview with Tom Wilder

TW = Tom Wilder
GAM = George Michael

GAM: We're in Graceville, Florida, right now. It's January 16th, 1995, and I'm going to interview Tom Wilder. Tom, why don't you start by telling us where you trained earlier and how you got to the Lab to begin with?

TW: I went to the University of Alabama, and got degrees in Engineering�an M.S. It was undesignated. My B.S. degree was in Mechanical Engineering. I took a lot of math courses, and was asked if I would teach algebra to incoming engineering students, and did for seven quarters. Then I taught a year at a junior college in Meridian, Mississippi, where my home was.

Then I went to work at what's now NASA, but in those days was known as NACA (National Advisory Committee on Aeronautics), at Langley Field, Virginia. My work was theoretical, with structures. We were called aeronautical research scientists. I spent four years there. Aberdeen Proving Ground and the NACA operation at Langley Field were two of the most advanced places at that time in computing. They both had those George Stibitz relay computers, big relay computers (BTL Computer�Model V).

We had a lot of problems in those days where we could essentially get the equations, but we didn't have the computing capability to get the answers. One of the problems that came up, for example, was a wing on a fighter plane that didn't have the strength it needed. It was a plane�in fact it was made by Northrop Aviation�that had two rocket pods, one on the end of each wing. And the performance of the plane was such that they finally wound up calling it a night fighter. They more or less based it in Alaska, because I don't think it had the general characteristics it needed to be an all-around interceptor.

So, during the Detroit Air Show, years ago, one of the plane's pods flew off and killed several people. I think ten or twelve people were killed when the pod struck down in the crowd that was watching the air show. It became a problem for Northrop to get the wing redesigned with enough strength. There were some ideas at the time on how to make a particular kind of wing that would be stronger and have a good strength-to-weight ratio. We had worked on a design at Langley Field, and Northrop engineers came out and wanted to use it.

I give this as example of the kind of computing we did: One of the scientists there at Langley Field went home for the weekend and knocked out the equations that would be used for designing the wing�all the parameters involved�and brought in the equations. And the boss called me in with the engineers from Northrop, and asked me if I would do the calculations to help them out. We normally wouldn't do anything to help just one particular contractor, but, because of the push on them to redesign the wing, we went ahead and decided to do their parameters first, to help them. They said that they had the Binac computer at that time, and so they would be able to solve it themselves. But, hey, they were kidding. It would have taken them six or eight months to do it, and we did it in a couple of weeks, you know? I don't think they had the capability out there�there were few programmers at that time in the country.

So, I made several trips back and forth to Washington, and I did the calculations on the old SEAC (Standards East Automatic Calculator), which was a forerunner of the Univac. It was an interim computer on that particular development. The first three Univacs, as you know, went to the Census Bureau, the basement of the Pentagon for the Air Force, and the Army Map Service. Those were units one, two, and three. And then four and five, of course, went to the Atomic Energy Commission�the fourth going to NYU and the fifth Univac to Livermore.

So, I had a pretty good reputation at the Bureau of Standards for solving problems, using the computer, and knowing what I was doing. They didn't know how long it took sometimes for me to understand things, but when I went up there I looked much brighter than I was. So, I think that when the need came up to solve the problems for the H-bomb feasibility and the hydrodynamics and all, that they were looking for anybody who might be able to help them. One of the questions you asked was who called me from the Lab. It was Robert Jastrow.

GAM: Jastrow? That's marvelous.

TW: Yes, Bob Jastrow called me and asked me if I would come out there and do it. And he said that they couldn't pay me as much as I was making at Langley Field. I said, "Well, that's not important. Pay my way across the country, and I'll come out there and help you." so I talked to Katrina about it, and she said that, yes, she understood that all of the kind of work that I was interested in was being done on the west coast with aircraft companies and all, but that she didn't want to stay out there over ten years because her father was living here, close to Graceville. And she didn't want to be out there as he got older and died.

GAM: Surely.

TW: So, she said, "How about let's agree just to go out there for ten years?" That was the agreement we made before we went out there. So, I went to Livermore. I don't recall any details that anybody said that I was going to head a group to do computing. I don't remember ever talking about anything other than just going out there and solving the problem, but the first thing I knew, somehow, I was flying back and forth to Philadelphia. And I don't know when you got there, but�

GAM: I got there in April of '53.

TW: Well, some of it was already over by then, because when I went there, it was November 4th of '52.

GAM: The Lab was two months old then, yes.

TW: Yes, and it seemed like there were only about a dozen guys around. I got to pick out an office along the old hall there. And they had two sizes. There was one that two guys could fit in and one was the toilet between two such rooms; but it was the wing of an old hospital barracks. And in the hospital they'd have a room and then a toilet that served two rooms, and then a room and a toilet and a room, and that was the pattern down the hall. So, I chose the toilet, because I figured I could be in there by myself and there wouldn't be room enough to get somebody in with me. I chose the toilet next to Teller's office, so he was on that hall at that time, and stayed there for, I don't know, a year or so.

So, Teller wanted me to go back to Philadelphia right away, and he also wanted me to design a tower for the shots. But I told him all the towers that needed designing had already been designed, and there wasn't any need to get into that. All they needed to do was get in touch with somebody in the steel fabrication business—PG&E, for example, could give him a tower that would work.

So, I flew back to Philadelphia. Of course, I had just come from the east coast all the way to the west coast, and then now I was going back to Philadelphia. I put Katrina and the kids in a motel, and we managed to buy a house—845 Princeton Way. And I started going back and forth to Philadelphia. Well, Art Biehl�and I don't know why Biehl wasn't in the Theoretical Division�maybe he didn't get along, you know?

GAM: Who knows?

TW: But I don't think he was ever in the Theoretical Division. But he was interested and had some problems to solve, so he gave me one. And we called it "Wild Biehl"�like Wild Bill Hickok, except�it needed a code name, I reckon. So I programmed it, flying back on the plane. And I put it on a computer when I got there and got the answers.

Well, if you're really careful, you don't have to do much debugging. And so I didn't have any problems.

GAM: Oh, how things have changed.

TW: You've got to think a little harder when you program them so you keep the bugs out.

But I had asked Biehl very carefully, "Now, is this all you're going to do?" "Yes." "Now, you won't want any more answers?" "No." "All right�these dimensions, these numbers, that's going to be complete, right?" Well, see, it wasn't a floating-point machine, so I scaled it and got all the answers. But, of course, when he got the answers back, he wanted more run�outside the range, of course. So, I had to redo that. He didn't understand, but I thought I'd been very careful so that I wouldn't have to go outside the scale I had in there. But I did.

The people who were hired as programmers at that time�Jules Mersel was one of them.

GAM: Yes, I remember him.

TW: And the per diem at Langley Field had been $8 a day. There, it was $10 a day for a room. And so what the kids had done was take the $10 a day and multiply it by 30 and get $300, and then two of them would room together. And they would rent an apartment in some new apartment houses�the Flamingo Apartments. This was in a run down section of town. But, that's where we would stay. They had four or five apartments.

GAM: This is in Philadelphia?

TW: In Philadelphia. All five Univacs were on the same floor in a Remington Rand plant there. And the first one had been used to predict the outcome of Eisenhower's election, and November 4th was election day.

So, I arrived in California on November 3d and spent the night in Bakersfield. Then I traveled on up to Livermore, and stopped at a filling station the first day to ask where the University of California Research Laboratory was. And the guy there said he'd never heard of it. The Radiation Lab, I think they called it�the University of California Radiation Lab. I said, "Do you know where the Radiation Laboratory is for the University of California?" and the guy in the filling station said, "No, what do they do?" I said, "Well, they do work," you know, and I knew there was a heavy classification. And I said, "They do work with atomic energy and stuff like that." He said, "Oh, you're looking for the atom bomb plant. It's right down at the end of the road." I said, "God, I thought this was all supposed to be classified!" I mean, the guy in the filling station knows they're making atom bombs down the road! But the problem was that the California Research Corporation was located there in that complex, building a boiling-water reactor or something�they probably were keeping that a little classified, too. So the people in town thought they were making bombs out there; so that was it.

And then at the end of Princeton Way�I suppose they had a "Harvard" and a "Yale" and all the ivy league schools named on the street signs in this little subdivision�the signs that they had put up at the end of the streets were stamped steel. The letters were embossed in steel. And the one at the end of the street had been taken and painted white, to get rid of the name, turned upside down, and a different street name put on it. But you could still read it. It said "Atomic Street," so apparently somebody had objected to naming it "Atomic Street," and changed it to one of the names like "Harvard" or something. I don't know whether that sign is still there or not.

Sterling Colgate lived right down there, right where that sign was, and Rosie Colgate. Rosie Colgate was a programmer in the early days. She had a baby and wanted to come back part time, but we weren't doing part-time work. If somebody worked half time, it would take them twice as long to get a job done. And we couldn't assign a job that took twice as long when we were working towards getting it done. But she didn't come back to work after she had the baby, but she had programmed up until the time she had the baby.

And they finally moved Univac number five out there to where we were. We had planned to go around the clock, because those computers actually work better if you leave them on 24 hours a day, rather than cut them on and off, because of bringing the power up on the filaments in the tubes.

And that was another problem we had, of course. We had to buy tubes, and we bought tubes by the case and tested them. And if they had the ranges that we wanted, we would select them. So, we selected them ourselves rather than buy tubes. Let's say we could buy tubes for a dollar, and if we had bought them from Remington Rand, they were $7 or $8 or $10 or $15 apiece. So we bought them wholesale, tested them, and didn't sell the rejects because that would have competed with somebody in private enterprise. So, it made it an expensive proposition to get those tubes for the Univac. We destroyed many new tubes.

I think Lou Nofrey was the guy in charge of the engineering group that provided the maintenance.

GAM: Right.

TW: Cecilia Larsen, I think, worked for that group

GAM: She worked for Lou, yes.

TW: Yes, she worked for Lou. So they kept the Univac up and kept it running around the clock.

One of the things I felt was worthwhile doing, and we did, and I think it's still being done, is that I set aside 10 percent of the time around the clock�for about seventeen hours a month, something like that�when we would let anybody work on anything they wanted to. So, if a guy had a project�I think Pennington was interested in a monorail train that was stabilized with some kind of centrifugal mass, or something. And there was a guy, like Louis Henyey, who was an astrophysicist. He'd come out and had time on it. John Gofman, the doctor who studied heart disease and all, came out and asked me to help him with some of the correlation on the data he had; they had about fifty thousand tests on blood that they put in a centrifuge and got the lipoproteins precipitated.

So, I think today they still probably have a policy of letting people who want to work on their own stuff, and have an interest in something else, have computer time. Anyway, there was some loose time.

GAM: I think that's great.

TW: And I think it helped keep people who would probably not have wanted to stay there if they hadn't had a chance to follow some of their own interests. That's the way it got started. And I guess that's why I came along at a good time, because they needed a lot of computing capability to essentially design the weapons. There weren't many people around at that time who had any experience, and knew the math or how to do it.

I remember Teller asking me, "What did you study so that you can do what you're doing, because we want to set it up at the University of California to have a supply of people that are coming that will know." And it was hard for me to describe to somebody what it is that I knew and had learned over the years�in particular four years at Langley Field�of solving problems which were partial differential equations a lot of times, and solving them using difference equations, and using what we call numerical mathematics.

At that time there were only a couple of books written on numerical mathematics, and I think both authors were still alive and young. Now, of course, it's a big field and everybody knows what it is. But at that time, there was hardly any way to set up what would have been a formal training course in how to learn the mathematics, the numerical differentiation, and everything else that you needed to know to do it. And so I don't think I was too much help in saying, here's how to make clones of these guys who know how to solve the problems on these kinds of calculators. You could see even then that some day they were going to have millions of programmers, you know?

GAM: I don't know that everybody had that vision, but I certainly agree with you.

TW: Yes, there was going to come a time when you couldn't possibly have enough. And we could see the software, which is one of the things you might be interested in. Grace Hopper was hired by Remington Rand back in Philadelphia, and she had a nice group of people there that were working on early compilers and languages. And I went back and visited her; usually if I was on the east coast I'd say hello to Grace and visit her. She had two very beautiful young girls there who were graduates, I believe, from Vassar. I think wherever Grace finished school, these girls were from the same school that Grace had attended. I said, "Grace, how about letting one of those girls go out to Livermore and give a series of lectures to our guys out there about these languages and what we can look for in the future and all?" And she did.

I didn't get to see the girl when she was out there. But she did come out, and stayed in a motel there in Livermore or Pleasanton. She dove in the pool and broke her front tooth off�hit the side of the pool. I felt bad, because she was a very beautiful young lady. I don't even remember her name, and never did see her after she broke her tooth, but I heard about it.

So, Grace told me what the idea of a compiler was, and I said, "Oh, yes, it's simple. I can understand that. That would be helpful to some people." Of course, when you're in the throes of getting something done like solving a particular problem with some particular equations, you're not interested in developing FORTRAN to do it. You just go ahead and do it, program it straight out without a language.

But I went back and, as you remember, we used to have clearance problems sometimes. Merritt Elmore was one of them that had a clearance problem. I think it had to do with his sister having been married at one time to a fellow who had some Communist connections. So, they had Elmore hung up, and he couldn't come inside. And, of course, we hired a bunch of people all along, and every time we hired somebody they were hung up until they could get cleared. So, they kind of sat outside the gate. I asked Elmore to work on this compiler, because it wasn't classified in any way at all. And so he wrote�

GAM: LMO?

TW: He wrote the L-M-O, or Elmo compiler. And it was the first compiler in the country.

GAM: It did do formatting?

TW: Yes, it took the raw data, just numbers. Then you formed columns with or without zero suppression, and all that stuff, and it numbered the pages if you wanted. It had some things like that in it as I remember. And I have a copy of it!

GAM : Oh, that's wonderful!

TW: And I'm not showing you this because Elmore said such nice things about me. I think he was a little embarrassed because I said, "Come on now, this is your job, you're doing it, it's yours, you put your name on it." That's why it's called LMO, for Elmore. [1]

GAM: That's great. Oh boy, we have to get a copy of this, Tom, and put it in our archives.

TW: Take that with you.

GAM: I can have it?

TW: Yes.

GAM: Okay. Thank you. That's wonderful.

TW: In fact, I'll just make a copy of the first page and keep it, because he put the little note there.

When Bob Price was hung up, not cleared, I asked him to do an inventory program, which would be a commercial application. It would keep up with all the parts and everything for Lou Nofrey and his operation.

GAM: I remember that. I was there in the cooler then. Everybody in the cooler had this three-word input thing. You came in and talked about the three-word input for this inventory. Every item had three Univac words, as I understood it.

TW: I've forgotten the details. I had a secretary�I can't even remember how any of this came about. I don't know why I had a secretary, but I remember I had a secretary named Pat Walsh, I believe, and I think she married while I was there. I don't know what her married name was.

GAM: I don't remember her name, but I remember her.

TW: Bob Price was married to Mary when he came. In fact, Bob Price came and said that his wife was pregnant and he needed a job. And I said, "What's your background?" And he said he had math, you know, and he'd gone to�I think he had some time at Duke and also at Georgia Tech. And Mary's daddy was the administrator for the Grady Hospital in Atlanta, which is not only a big job, but a big hospital. So I said, well, sure. He seemed well qualified and a bright, young fellow, so I took him on. I figured he'd do real well.

GAM: He did.

TW: And a lot of times, George, in a project like that, you've got to get some real bright guys. And they may not stay with you forever�they may do something else�but you've got to get what you can while they will agree to stay.

Let me tell you an anecdote. Teller was back here at Troy University. He had some kind of arrangement with Troy University, which is not many miles from here, up in Alabama. And he would come back and spend two or three days with them each year on some kind of honorarium or something. So, one time he was up there, one of the physicists there asked me if I would come and have lunch, and said Teller was going to be there.

Well, there was a very wonderful young girl who worked for us here, and she went to Troy. So she was up there at Troy at the time, and I knew her father and her mother real well. She had been Miss Peanut or whatever we call it when they're the Queen of the Peanut Festival. So, I said to her, "I'm coming up there to visit Dr. Teller, and while I'm up there, I'll get together with you, and we'll go listen to Teller," because he was going to have some kind of lecture or something.

Teller played the piano while he was there. So, we renewed our acquaintance. "Edward," I said, "I'd like to tell you an anecdote about you sometime." And he said, "Well, tell me now." So, they had a little microphone on him and a recorder or a transmitter, and they were keeping up with everything he did for a couple of days, because he was going to be "Man of the Week" on an ABC news program.

The anecdote goes like this: Teller asked me to solve a particular problem, and I chose to do it using Monte Carlo techniques. That involved developing some random direction cosines, which I imagine they're still doing out there. And Elmore, I think, got in on this, too, eventually. He wrote subroutines to develop these random direction cosines. So, there were no trigonometric functions on calculators in those days. You had to evaluate them with a power series. The Univac had nothing like cosine or sine on it, so we just ran a power series out, and let it converge, and we'd use the number that way.

But, you remember, the Univac had a buffer on it, so you could bring in 60 words at a time, in a kind of parallel processing, and that didn't take any computer time. The 60 words would transfer into this buffer, and you could still be computing at the same time. So, what I did was run out a hundred thousand sets of these on tape and just let the tape feed them to us, rather than keep generating these cosines and expanding these series, because it was too time consuming.

So I did it that way. But it wasn't anything to be particularly proud of, and it irritated me that I had to expand these series to get these direction cosines.

I remember explaining it to Bob Price in my little toilet office on the door, where I drew a ray on the door, you know, from the bottom corner. And, see, I could swing the thing around on a plane. You know what I was doing�I was explaining to him why you needed random direction cosines, because you can't do it�you'll get a bunching at the poles, you see, if you don't do it right. You have to have a cosine function in there.

GAM: Yes.

TW: We sat down there on the floor, and I was explaining it to Bob Price, because I always felt like anything like this that I did initially I was going to have to have somebody to keep doing it or understand it. And then I'd get the next new problem that needed solving. So, Teller stopped by one day, and he said, "How are you getting along with that problem?" And I said, "Great." It was about a week later. I said, "Great. I've got it solved. I can get the answers, but I'm not proud of the way I'm developing these random direction cosines."

So, he said, "Well, Johnny von Neumann told me a way to do that, but I don't remember exactly what he told me to do. But he showed me a way of doing that." So, that was the end of it. He left.

Every Monday morning, Harold Brown, who was in charge of large weapons, Johnny Foster, who was in charge of small weapons�that's about all the organization we had at that time�and I�we would meet on Monday morning. And Johnny and Harold would say, well, we need somebody to work on this, and we need to get some physicist to think about this or that problem. Meanwhile, in order to meet the time schedule and restraints we had, we needed more time spent on this one, and it's more important to get these answers than those. So we kind of decided who was going to get time, and who was going to spend energy on solving different problems, so that we could get everything done together.

On this particular Monday morning, we met, and Harold said, "What's wrong with Edward?" And I said, "What do you mean?" "Well, he's preoccupied with something, and we can't�it's hard to talk to him." And Johnny Foster seconded it, "Yes, we get in a meeting, and he's thinking about something else." I said, "Well, maybe I know what it is."

So, after Johnny and Harold left, I went in Edward's office, and just as soon as I opened the door�you know how tense he is�he said, "I haven't been able to figure out yet how to do those random direction cosines!" So, I stood in front of his desk, and he sat there, and I said, "Well, why don't you call Johnny?" You see, all he had to do was just ask him over the phone, you know? So, he called Johnny at the Institute for Advanced Study at Princeton. He had his number, so he dialed him and got Johnny on the phone. He said, "Johnny, fly out here. I want to ask you something." You see, I figured he'd just ask him on the phone. There was no need for Johnny to fly across the country. Or, Johnny could have mailed us something. But, of course, I'm sure that Teller wanted a visit from Johnny for many other reasons. And Johnny was retained by us and everybody else in the country at that time, including IBM.

GAM: True.

TW: So, Johnny flew out, and I went in Teller's office in a couple of days, and Johnny explained to me on the board how to find random direction cosines. And he was a true genius. He not only had a genius for math, and understanding it, but he had a genius for tutorial activity. He knew how to teach me. He could look in my eyes, and�

GAM: Track the light.

TW: I'm too dumb to be anything but honest, so he could look in my eyes and know I wasn't getting it, or know I was.

GAM: Marvelous!

TW: So, he knew exactly how fast to go, and it only took him a few minutes. So, then, I'm sure Elmore wrote that up, in some subroutines, so we could generate them.

Now, since then I've seen that in print, in a book on Monte Carlo methods. I've got it up there. And, of course, I've got the original that von Neumann showed me. I wrote up some notes on it, too, of how to do it�an ingenious little method.

GAM: That's great.

TW: Yes. Just a few nights ago, the television station up here in Dothan, twenty miles from here, was bought out by a group. And one of the names in the directors of the group buying it out was Marina von N. Whitman, who is Johnny von Neumann's daughter. So, I had seen her picture earlier in the Annual Report of the old Manufacturer's Hanover Trust Bank in New York. We corresponded with them, and they furnished us with our group insurance and things like that.

So I wrote her, and I told her that her father had taught me how to do random direction cosines. And I told her the little anecdote about her father flying out. She wrote back a nice letter and said she appreciated my telling her this, because she was unable, of course, to appreciate her father's genius except when people like me would tell her that he was a genius and what he had done.

GAM: That's great.

TW: Manufacturer's Hanover merged with Chemical. And she was Chief Economist for General Motors�so no doubt a very talented lady. I told Teller the anecdote at Troy. And, of course, I don't know that he even then appreciated the fact that he should have just asked Johnny on the phone and not have Johnny fly across the country to tell him. But at least it gave me a chance to evaluate the genius of von Neumann, and he truly was a genius.

GAM: This adventure with Teller must have been in late '53?

TW: Yes, because, you see, I stayed there only about eighteen months, and I came in November of '52. So that would have taken me into early '54, when I went down to Convair.

There was a boy there�I can't remember his name now�who left before I did. I reckon he was probably the first one in the group to leave, and he went down to North American, at Downey, to work on what they called the Navajo Project. And after he got down there, he told them, I reckon, that they ought to contact me, and see if I wouldn't come down there and help them. I wasn't particularly looking for a job or wanting to do anything. I liked what I was doing, but they asked me to come down. And so I went down there and talked to them. The project wasn't all that interesting, but they had a bunch of people. When they canceled the project, I think they laid off ten thousand of them in one day!

But, I think because it stirred them up, then they probably asked some people at the Bureau of Standards, and I could probably figure out who. Also, I know, for example, von Neumann had a contract. He was retained by General Dynamics, a Convair Division. I think he advised Convair to stick with the development of what followed from the X-2.

GAM: The rocket airplane.

TW: The X-1 was a ramjet. The X-2 was the ICBM type. So, stick with this, because he knew the H-bomb was coming through, see? You didn't have enough accuracy to really hit a target, but you didn't need to hit it too close if you had the H-bomb. So, he said, stick with it; he could see what was coming. And he told them to put your own money in it and stick with it. And to this day they are still firing Atlas missiles! God�how many years ago was that?

GAM: In the '40s.

TW: In the '50s�forty years ago. You've got it right now, George!

GAM: It was a long time, I'm telling you. Let's see, Sid came in September of '52 also, so he must have been around when you�

TW: Sid was around the whole time. But Sid didn't seem to find himself, exactly. He didn't seem to do any physics. I'm sure he'd done a dissertation or something, but he didn't have any interest. He never worked in anything that I knew of, but he was hanging around. He seemed like a gentle soul, and interested. He seemed very interested in what we were doing, and kind of wanted to get into the computers to some extent, but I didn't know what. He didn't seem to know math, and he didn't seem to know computers, and didn't seem to know physics, but he was there.

There were some other people that were there, and it was hard to understand why. Now, Leland Cunningham was a very capable man, and, oh, what a sweet gentleman. He was there, but I never asked him to do anything. He was doing his own thing, and should have been. Of course, before the digital computers came along�like Binac, and Univac, and so on�the only people I knew of in the whole world that were doing any computations were the celestial mechanics. And Cunningham was one, and a good one. So, if you had a problem in forward integration or something like that, you would have had the right guy.

If you've got time and enough tape I'll tell you another anecdote that happened later.

GAM: I have more tape.

TW: When I got to Convair, of course, they were getting the first 1103. They had the largest analog system in the world, and they wanted to tie the two together. And I'll tell something about that later. One of the things they wanted to do was take the missile, the Atlas missile, and fire it, and guide it with some inertial guidance in the first phase. Then, when it went ballistic, and no more forces acting, no more thrust, it would hit the target. So, there was some computation to do, and some simulation of internal, on-board computer-like devices that would direct the Vernier rocket, the little rocket that is fired out the sides.

So, Convair had gone through good times and bad times and all. You can easily develop a theory that if an aircraft company expands rapidly, and contracts, and then expands and contracts, that the guys who stay around are the least capable. So, some of them that were down there weren't too impressive, and some of them were. But, in general, they weren't very able to get a job done very well.

So, the engineer, a kind of officious type, came down and said, "Here is the initial input for this problem, and we want you to calculate the trajectory and all that�like all along�and print out all the parameters and everything." So, it was duck soup�there wasn't anything to it. So I did it, and put it on the computer, and got the answers.

And the guy came down and said, "You've got a mistake." And I looked at him�I didn't say anything ugly, like "dumb ass" or something. I just said, "How do you know there are mistakes in it?" How would he know, you know? So, he said, "Oh, we had another guy do it. We had a consultant run it out, and he got different answers from you." I said, "Well, how in the hell do know that he's not wrong, and why do you think I'm wrong? Why don't you think he's wrong?" Oh, he was�what was the guy's name? He was the head of the Cincinnati Observatory for computing orbits for asteroids.

So, I said, "Well, why don't you call the guy, and tell him to come out here, and we'll get together, and I'll tell him where his mistake is." So, the guy did. He came out�a very nice, nice guy. I don't know whether it was Brady or one of them up there at the Lab who was always interested in calculating the position of the twelfth moon of Jupiter. And we'd say, "Hey, we're going to do it�next week we'll get together, and we're going to find some time, and we're going to forward integrate it and locate it."

GAM: That was Joe Brady.

TW: Joe Brady, yes. I never did get around to doing it, but this guy did it. I forget his name. Anyway, he beat us, because he had access to a computer, and he cranked it out, and so we never did do it. But, anyway, that's the guy, and so he came out. He headed an observatory in Cincinnati. In fact, he discovered an asteroid and named it after his daughter. They name them like�"Carolyn 1734" or something, or Marilyn or whatever her name was, Josephine or something. [2]

So, he came out, and I pointed out to him that it was�you know, sometimes you get like seven points�zero point, three points before and three points after�and you start this forward integration by having these differences that you develop. And so he had developed these differences using seven points. And I said, "Well, what about these three points here before time zero? You know these little rockets cut on here at time zero. What did you use in here before time started for these things?"

And so, you see, his mistake was that he had assumed that there was a continuity across time zero. And there wasn't, of course. And the numbers over here�since you're starting something with acceleration�the numbers over here were huge, and negative, and in the wrong direction and everything. It made enough of a mistake so that as he went on out, he was off from my answers. I'm sure they were close, approximately, but he saw it quickly when I told him. Of course, he had written the book on forward integration, and starting out with these differences and all. But it works differently for a planet or an asteroid that has no forces acting on it other than gravity. So, he wasn't used to something that cut an engine on right in the middle of an integration!

Bob Price, Jack Rose, and Bob Gerkin went with me when I left Livermore. And I'm sure that people at Livermore were a little upset that we would leave. There are people who think that you ought to be loyal, like George Michael or somebody else, and stay with them, because it was a nice place and nice people, and enjoyable. But, I tried to explain that I love to design bridges, but I didn't like to stay and take the toll up after I got the bridge built�so my job was different. And once I had been there eighteen months and solved all the problems, and we had them set up, and had people like Shirley Campbell, I was ready to move on. Shirley was a wonderful person.

GAM: Shirley has retired, but I don't really know for sure.

TW: She was good.

GAM: And very quiet and competent.

TW: She got married, and raised chinchilla rabbits, and had a baby, and stayed out a day and a half when she had the baby. I mean, she missed a day and a half of work. She was very conscientious. We went back to the Statler Hotel in Washington�that's where we stayed when were in Washington. So, she went back�I don't know, we stayed a week or two for some reason. Oh, I imagine that we were using the�

GAM: NORC (naval ordnance research calculator).

TW: No, we were using the computer in the Pentagon, and we were using one at the Army Map Services.

GAM: Oh, those, yes.

TW: On all shifts. We'd go back and take the night shifts and all, and run our work when they weren't using them. So, we had gone back. And the guy who managed the hotel had graduated from the University of California at Berkeley. So, he still had some friends there. And I imagine that's why somebody in the Travel Office was always making us reservations there�"and this is the place to stay when you go to Washington."

Now, there was a good reason for it, because at that time, you know, the Korean War was on, and things were booming here. And they had a three-day rule that you could only stay three days. You'd go in and say, "I want a room in the hotel." And they'd say, "Fine, but we have a three-day limit. You have to get out and then let somebody else have the room." But they waived that rule for us from the Rad Lab. And so we could stay there a week, two weeks, three weeks.

Shirley got a room. I got a room. If there were a couple of more in our group, they had rooms in the Statler. And it was pretty good living�it was more than $8 or $10 a day. Shirley stayed there a night or two; then she called me and said that she couldn't sleep because the room cost so much. I know the feeling. I've been in rooms that cost me more to get three hours' sleep than I made all week.

So, I went to the manager, and he was really nice. He said, "Yes, we've got a policy. We have people like this all the time. We have a policy of setting aside a number of rooms (I think they had six of them) that we let people have for a real low rate, if they're bothered in a sense about how much it costs." So, we moved Shirley into one of those, and she was happy and was able to sleep.

One day, Biehl, Mainhardt, and I went back to Washington. I don't know who the other guy was that was with us. I don't know whether he was a mathematician or a physicist or a mathematician-physicist�I've forgotten who he was. But we got in on a Sunday afternoon and went to the Statler Hotel. We cut on the TV, and the news reported that a locomotive for a train with no brakes and had knocked out the barrier at the end, and had gone into the lobby of a Union station and fallen through into the baggage room down below. And what a mess! And it had happened just right then.

We had the car down in the garage below, so we went down and jumped in the car, and ran around to the station, and got out and went up there. And it had just happened�the dust and smoke were coming around.

So, we were on the outside. And we wanted to know, well, where in there is there something going on? So, Biehl�a real bright guy�said, "Boost me up to the window, Tom." They had some little windows up at the top, like that. He said, "Boost me up and let me see." So I backed up and put my hand down, and he stepped on my hand, and I pushed him up so he looked in the window. I said, "What's going on in there?" He said, "I don't know. We're looking in the ladies' room!" He was. So, we got down and went around, and we had a chance to see. It was chaos. It was bad.

GAM: Quite an event.

TW: We were standing in the lobby at one time. I don't know whether it was the same trip or not. And I imagine you remember Mainhardt, don't you?

GAM: Oh, yes.

TW: He was very much of a sharpie type, you know, and alert.

GAM: An operator.

TW: "Operator" is the right word. There are other words to describe him, too. I used some of them with him, and I shouldn't have. I shouldn't have talked to him so ugly. But he was bad; he was bad news for an operation like that.

Anyway, we were standing in the lobby, probably trying to check in or out. But at certain hours in hotels, you know, they just�oh, man, what a mob! Everybody's trying to get their bill paid or something. And you're just all standing around, and you can't do anything.

Mainhardt leaned over and whispered to Biehl and me, "Don't look now, but there's Walter Ruther!" "Walter WHO?" said Biehl. Biehl is not devious. "Walter WHO? Where? What?" He wasn't two feet from him, you know. And Walter Ruther, I think, was the one, or his brother is the one, who had gotten his arm blown off with a shotgun when they tried to assassinate him, essentially. And he had some huge bodyguards, about three of them, standing around. And he was a little, tiny fellow with red hair. But he had some big bodyguards. So, when Biehl blurts out, "Walter Ruther�WHO IS THAT?" we got some attention! Biehl was a joy, eh?

As I say, I don't remember exactly what�we didn't have positions at first, you know. If they had somebody who wanted to do something�I think they had a refrigeration expert from the local repair shop downtown working out there, because they were working with cryogenics or something, you know what I'm saying? I'm exaggerating, but people were there, and if they didn't fit in, nobody wanted to chase them off. They let them hang around. It's nice, but it's unusual to try to get a job done that way. I think you've got about 8,000 of them out there now, doing the jobs.

GAM: That's what I'm beginning to believe.

TW: Biehl loved my basket-weave fence. It was hard, you know, because it was like rock. You had to take an iron pole and sharpen it, just chip out a hole for a fence post, and then weave those redwood basket-weave fences. And, God, I was so proud of mine. It was so beautiful. But it gave me something to do on the weekend, you know, when I was off.

And Biehl said, "I want one!" He lived a couple of streets over from me. Herb York lived there, too, in the same area, and Biehl. I think his wife was named Margaret. And Biehl said, "I'm going to build myself one."

So, a few days later I was over at Biehl's house, and he said, "Come on, I want to show you my fence!" Well, his fence went up and down like this, you know? It followed the ground. So he said to me, "How did you get yours so level?" I said, "Well, Art, you start at the top." He hadn't thought of that. He had started at the bottom, you see. And that's the way you are�you get things on the front of your mind, and you don't think about a little thing like how to keep the fence line level.

GAM: I understand.

TW: He and Mainhardt were always trying to think of something. Mainhardt had a proposition for everybody: If you think of a bright idea, let me have it so I can make some money. Biehl had a class at the University of California in reactor design. He had his class, his graduate students, design a reactor. So he had the design all laid out. And then he and Mainhardt were going to sell it, because they figured they could make it for $75,000 to $100,000.

Meanwhile, the government, Congress, in all their wisdom, had agreed that for any university who wanted a reactor, that the federal government would put $500,000 or maybe $1 million against the cost worldwide�any university worldwide. So, they said, hey, we can make this thing for $75,000 to $100,000 or something, and the government will pay for it. We will charge $100,000 or whatever. But, anyway, they had themselves something that would do.

So, I was up there in Livermore. Although this happened after I left, I still kept my badge current up there, and I'd go back and visit. And I reckon I could still do it�I don't know. I went up and visited with Biehl and Mainhardt, and I told them, hey, we got a bright idea about building this thing. They said, "but we don't know�how do you contact somebody to do it?" I said, "Well, let's make a list of prospects." So we came up with 38 prospects that may be interested in furnishing the money and so forth. And one f them, of course, was Howard Hughes�why not? We even tried to call him. And we got some guy, like a right-hand man for Howard Hughes. I met some of his right-hand men later on, like�

GAM: Mayheu.

TW: Yes, Robert Mayheu was one, and Noah Deitrich was one. But this was a guy, Dell Webb, who had a construction company. And I think he owned part of the Yankees at one time or something. I think the company name and his name were the same; I think it was on the New York Stock Exchange. Anyway, when we tried to call Howard Hughes, we got him.

But, they finally wound up with Aerojet General, and then they formed Aerojet General Nucleonics. And I visited them out at Walnut Creek when Art Biehl was out there with his graduate students. He had a little office up over a store or something. But, apparently, they were successful. I don't know what ever happened.

I ran across Mainhardt again. Later, many years later, I would go stay about a week at the Pentagon. Once a month, I'd spend a week like at the Pentagon. And they let me look over, like for the DDR&E (Director of Defense Research and Engineering) and all, the research projects and the contract they had. One thing they were proud of was their list of the twenty most pressing things to develop. So, they would say, "Look these over." These were Top Secret, of course. "Look these over and let us know what you think about these twenty things." Some of them were not particularly reasonable to do, but they were bright ideas, you know.

So, just going through the drawers up there one day, in the files, I ran across Mainhardt's name, and looked in there, and he had these little jet bullets, you know. Do you know what I'm talking about?

GAM: Yes, I know about them.

TW: Well, I saw that back there in the drawer, and� He had a contract, apparently, to make them or develop them or something. I don't remember if anything came of this, but several other ideas were used.

GAM: Many, many were used. Did you have anything to do with the LARC?

TW: Yes. When the 1103 came in�well, let me go back. It was pretty clear after a few months' operation that you certainly wanted something bigger, faster, and all of that. And, of course, developments were going on all over the country. You've always got the IBM "stooges," I call them, who want to get IBM, and rightfully so. IBM is a very big company. And they've been good to me, although I've never worked for them. But they've always been good to me in owning some stock or thinking that they were going to do well. And they did for a while.

At the Lab, they wanted to come in and sell to us, naturally, because we had a Remington Rand Univac. Remington Rand, of course, was way out in front of IBM at that time, and there were a lot of reasons for that. But they said, "Well now, we offer a 701." Well, there wasn't any reason to buy a 701, because it wasn't going to do much more than the Univac, and why split up and have people having to know both or do something else? We could just go ahead and use the Univac.

Well, as the pressure began to build to get a computer that had more capability than the Univac, the question was, is Remington Rand going to build us something or will we work with two different manufacturers? And there are advantages and all. So, Cuthbert Hurd worked for IBM. And he came out, and he began to try to tell us that we should get a 701. And we said, well, there's something you can do�at least make it better, like the drums. Lace them in such a way that we get more access, or bigger drums, or do something. So they came out with what they called a 701A. And then we said, well, you know, you really ought to be able to do better than this.

So they came out with a 701B. And I think we ordered one or two, but by the time they went back and did the engineering, they decided they were going to call it a 704. This is not a rework of the 701, they claimed, this is really a new computer�the 704. So I think by the time we got it, it was 704, and it was a different thing.

But I began to�I was practically gone at that time. And I didn't like the selling, and I didn't like everybody trying to get in the damn act, or saying what to get, and having an opinion when they didn't know�

GAM: But we had 701s, you know.

TW: Yes, I think we did.

GAM: We had several of them. I remember I could never get my problems to run reliably on the thing.

TW: Yes, I think I remember them.

GAM: It had the Williams Tube memory in them.

TW: Now, wait a minute, wait a minute.

GAM: Yes. Well, in fact, in that brochure about the memorial for Fernbach that I sent you, there's a listing of when these computers showed up.

TW: Yes. I remember briefly, vaguely about it. But now, the LARC came up. The 1103 also had a cathode ray tube memory. The IBM 701 came in 6/54�so that was June. That was after I left. And we got two of them. And then we got a 704, yes. (The 701 was probably delivered on an "interim" deal to replace the CPC�then the 701 was to be replaced with a 704.)

Well, I probably wouldn't have fooled with the 701s. I don't think it was worthwhile.

GAM: I agree with you. I think we should have gone to the 1103.

TW: Yes, but only if they had kept going with something better and more scientific.

I went down with the 1103, and when we tied the big analog computer together with it, the guy that�Ben Ferber was his name.

GAM: I knew him.

TW: He said, "We've got to have some output, and there's a guy over here that makes a cathode ray tube that has letters on it. The guy's name was MacKenzie, or something, and it was a Sage system display that was for the North American Air Defense Command. They had a big, round, cathode ray tube. And they had a disk in the throat, and two yokes, so that you could select one of, say, 64 characters and form the electron beam. And then the next yoke would put it on the tube where you'd want it at 512 lines, or whatever, in position. And that's the way those things still work. So, he sent me over to the engineer, and I talked to this guy, a fellow that worked there, and I said, "We want one, so let's get it hooked up."

So we got the engineers from St. Paul together with him, and we finally got a Polaroid camera and left the lens open, and we would write out on the screen what we wanted to. Well, then, see, I was still running back and forth to Livermore every now and then, and visiting up there, and I told them, as the LARC was being designed. Did you know that there was a decision made that had nothing to do with the Laboratory or anything else that they were going to build a LARC and a Stretch? They wanted to give each manufacturer a go-ahead to build some bigger computers. So, the money happened to come through Livermore to do the next generation for Remington Rand, and the money came from Los Alamos (the Stretch) to do the next generation for IBM. And it was keeping them both in business building the next generation. Guys like von Neumann were in on those kinds of decisions.

So, I talked him into that for the LARC, and they put two of them on the LARC. Now, we paid about $64,000, as I remember, at Convair, and Remington Rand charged Livermore $400,000 for the two.

GAM: We had to completely rebuild them in order to make them work right. But they came to us from Stromberg-Carlson. They were called electronic page recorders.

Now, who had made the decision that they were going to build the LARC and Stretch independently of the labs? Was this a DOD kind of thinking?

TW: Well, either that or somebody. Of course, NSA was computer-happy. And I think it was just that it was a level in the government that said, "These computers are great." You know, Vannevar Bush said they'd never need more than five of them! [3]

GAM: Yes.

TW: But, I think he was wrong.

GAM: Apparently.

TW: But after they got the Univacs out there�and they first ordered three�then they began to make more. And finally they had 17 or 20 of them built before they really changed the design. And I think that they were saying, "Well, we need to go a generation beyond," because I think they could see a faster thing. And, of course, since the Univac, we've had the transistors, we've had the magnetic memories, we've had everything.

GAM: Well, integrated circuits now.

TW: The SWAC was the one with the Williams Tube memory on the west coast. And on the east coast�

GAM: It was SEAC.

TW: And it had a Mercury delay line.

GAM: Yes, I remember those machines with great fondness. They were pleasant machines�friendly, you might say.

TW: The Univac was�

GAM: The Univac spoiled Livermore, you know. It was a thoroughly checked machine.

TW: It was so easy.

GAM: And that made it quite a bit of difference to people who were just interested in getting results. But it was quite slow.

TW: Yes, it was slow, relatively speaking.

GAM: Well, fundamentally it was slower because it was a serial machine as opposed to the machines that now run parallel.

TW: Also, that Bell relay computer at Langley�I checked that out because I did a lot of my own calculations, and enough of it so that I could put one on the Bell relay computer, and then I could check myself against it. And it was five times as fast as I was. That's the thing�not very fast. But, it didn't stop to drink Coca Colas, and it didn't make mistakes, and it would run unattended and cut itself off in the middle of the night. I mean, you know, if you were running a problem and the end of the shift came, you could go home. It was an interesting machine.

GAM: I'd like to go back a little bit and probe more about any interactions there might have been between you and Sid, you know, because he ended up as head of Computation.

TW: Yes, and let me say something that probably ought not to be said: Sid was difficult to work with.

GAM: You ain't the first to notice that!

TW: Well, he was a pain! He was always sticking his head in the door and asking you a question about something, or butting in, and you could see what he was doing. And I didn't have time to play games with him. Sid wanted to run the entire computer operation, but such a job didn't exist then. When I left the Lab, some people were not too happy because I left, and certainly they should have been unhappy that three guys went with me, although I really didn't have anything to do with it. I mean, it wasn't my decision.

I more or less told Teller that you've got a guy here that wants to do it, you know? Let him have it and be in charge of it. He said, "We don't want you to go." I said, "Well, hey, you can get along without me. There's a guy around here who wants to replace me anyway�Sid." And I don't think his mind worked in such a way that he would want to get in there and find out what was going on and do it enough to learn how to make a decision, like whether to go this way or that way. And I'm sure that he listened, but I don't know that he had the experience or the understanding to properly guide it the way it should have been. And I don't think he probably could lead the young guys, you know.

GAM: At first he was all right, Tom. He listened very well. Do you remember Chuck Leith?

TW: Yes.

GAM: Well, he would listen to Chuck, and take his advice, and other's advice as well. And we did do a lot of things that were sort of heads-up because of that. Then he stopped listening.

TW: Yes. But the Laboratory was very young�it didn't have a lot of seasoned managers. I probably saw E. O. Lawrence twice the whole time I was there. I don't know why they named the Lab after him. I never did understand. I'm sure he played a role in getting the separate laboratory for Teller. I don't guess it's any secret that there were a lot of people at Los Alamos that were unhappy with Teller. I don't think Teller liked some of them, and some of them didn't like Teller. I think there was some anger involved, and so Teller more or less left Los Alamos.

There's an interesting anecdote I'll tell you. When Teller was pushing for his own project to do research in thermonuclear devices, there was a lot of opposition to it, like from Oppenheimer. "Oppy" they called him. And he was revered. He was an icon. They had reverence for him. Louis Ridenour was made Chief Scientist for the Air Force. The first Chief Scientist they had. And he was a very capable guy. He had some problems, but he was very, very sharp, very capable. He could get things done, knew how to go get things done.

The Strategic Air Command (SAC), as being part of the Air Force, and the Air Force, had been drawn from the old Army Air Corps, which was drawn out of the Army. Now, you've got the cream of the crop from the Army and you made the Air Force out of them. You got the cream of the crop in the Air Force, and probably put them in SAC, you see? SAC had some very, very good officers.

I knew Curt LeMay, the head of SAC after World War II. He just passed away recently. I initiated him into Tau Beta Pi at Ohio State University. He and I were both Theta Taus, which is an engineering fraternity, but he was also initiated into Tau Beta Pi. And I was the guy to lead him around and get him initiated into Tau Beta Pi, induct him up there at Ohio State. He graduated from Ohio State.

But these guys were very aggressive, and they felt like, "I'm in charge of defending the world, and I've got to keep these bombers flying, and I've got to develop these weapons." And they wanted the best weapons and the best planes, and the best guys!

So, when Teller was shopping for somebody to do it, and go ahead and back him on it�and he had a couple of congressmen lined up, and other people�I think Louis Ridenour more or less was a go-between for him to be able to say, "If they won't let me do this at Los Alamos, the Strategic Air Command will furnish me a laboratory, and I'll do it for them." Well, of course, now, that was going to get out from under the civilian control of the Atomic Energy Commission. So, no, we couldn't have that! And I think they capitulated, and I think that's why Teller got the Lab, probably was because Louis Ridenour was on this thing, saying, "Hey, SAC will furnish the money for you to do this. We'll give you a lab."

Well, I got to know Louis real well later. We used to go riding in his Mercedes Gull Wing�you know, it opened up on the side.

GAM: Oh, yes.

TW: We used to go out and have lunch at the little old town, Woodside, where they won't let anybody build anything that looks new. I mean, it looked like an old cowboy town. It's close to where I lived, close to Lockheed Research where I wound up. Louis was there. He headed Lockheed Research at one time.

But, anyway, back to '53, probably, Louis showed up at the Lab. A lot of people showed up at the Lab. They were looking for work, or looking for contracts, or just looking�you know what I'm talking about. There's nothing wrong with it. It's the way it ought to work. So Louis came. And I think at that time Louis had formed a little corporation or something�you know, a research thing to get contracts with and make money.

GAM: Sure.

TW: So, he wanted some work. He wanted us to give him some. Well, I didn't know at the time that Teller essentially owed Louis a favor. Now, this might not be fair, because I don't know all the facts, but Teller asked me to go to lunch with him and Louis. And we went down and�do you remember the old Chinese restaurant? It was about the only place in Livermore to eat in those days.

GAM: It still is!

TW: So, we went to the Chinese restaurant, and sat there in a booth. And Louis was saying, "I can do this. I can do that. I can help you with this. I can take a contract to do this. I�" you know? And he probably could have, better than any of them. But we weren't in a position to take on somebody from the outside or funnel out any work, do you know what I mean?

GAM: Yes.

TW: That would never have worked. And so I was turning him off at every step of the way! "No, I don't think that we can use that. I don't think�no, we can think about that, and we will, but no, no." I was not positive on it.

Now, had I known that Teller owed him a favor, things might have been different. I had great admiration for Teller, for many reasons, even though maybe somebody could see some faults sometime. But still, you know, hey, my father had faults. I've got faults. Teller had faults. But I liked him. And I got along with him, and I didn't have any problems with him, ever. And so, other people might have. I didn't have any problems with him.

But I didn't back him up, and I don't think he wanted me to. I think that he was saying, well, I owe this guy something. But maybe it was, Teller felt like, well, deep down in his heart, "Louis you could have done more for me than you did," and therefore he was cold to him. I didn't get any messages one way or the other, I was just saying, "No, no, no, no, no!" And I didn't know at that time that he owed him a favor.

But Louis and I got to know each other real well later on. There are some anecdotes there that I couldn't tell you. But I'll tell you, Fred O'Green, who became Chairman of the Board of Litton Industries, was my boss at Lockheed Missiles and Space Division. And he could tell some real good stories about Louis and Teller and Washington, D.C., and meeting in the Pentagon and all, and some of the decisions that were made later.

Fred O'Green is a very capable guy.

GAM: So, what was the upshot of your Chinese lunch, then?

TW: We just turned him off all the way around and never did anything with him.

Now, I did something one time: Dick Courant was also, you know, not only one of the �migr�s that came over from Europe, but he was a friend of Teller's and a great mathematician. He was well respected, and the nicest little guy you'd ever meet�a little, tiny fellow.

GAM: Yes, I remember him.

TW: Yes, I mean, he was so nice. And Teller told me to do something nice for Dick. And I get along. I mean, I know how the world works. So, I did something nice for him. You've got a Univac and we've got a Univac. You've only got eight servo mechanisms on yours, and we've got ten. I'm going to buy you two more. So we bought him two more. That's $200,000-worth of servo mechanisms for it. So, I justified it on the basis that we used ten, and we programmed for ten.

Actually, I suggested programming for nine to allow one down. There are usually eight, but, you know, certainly you wouldn't program for ten. We wouldn't operate it 40 percent of the time if we required ten! But, I'd say, "Program for eight, and then, if you have to, use one sparingly and maybe be able to get around it if only eight are up." So, we justified two more for him because, of course, if he had only eight, and one of them was down, nothing we had would run, probably, you know. We had guys�and you remember the old, funny operating deal we had?

GAM: Yes.

TW: The Univac is difficult to operate, and Lou Nofrey wanted to furnish operators. He didn't want us to hire the operator. I didn't want to fool with operating anyway. If you furnish me with guys, fine. But we had some differences in philosophy about how bright the guys ought to be.

GAM: Well, they ought to be very dexterous, and they certainly were.

TW: Right, but some of them took a long time to train, and a lot of aggravation and mistakes before you got them shaken down.

GAM: Well, yes.

TW: You know, we had these guys going back to NYU and the Census Bureau, and the Pentagon. And we'd get time on those computers and run problems everywhere there was a Univac; we were out there using it. There was one guy at the Army Map Service�I can't remember his name�a nice, young pleasant fellow. In fact, he was such a good operator�he was their operator�but then we'd turn our problems over to him, and he could do them for us at night, and would. So he helped us. In fact, usually if we had an operator that was sent back with our programs and problems, then we would get their operator to work with us, to be there, because we didn't know our way around, whom to call for problems, or how to get things done. So we usually went to this guy, who was very good. Hayes�his name was Hayes.

GAM: Oh, yes. I remember him.

TW: Jim, or something like that. Anyway, one day I was at the Sands in Las Vegas, out at the swimming pool, and his uncle, Gabby Hayes, was sitting there with three beautiful young starlets. So, I said, Mr. Hayes, I know your nephew in Washington, D.C. "Yes, how's he getting along?" A very, very pleasant fellow, Gabby Hayes was. But that was a Hayes guy.

GAM: It wasn't Cecil, was it?

TW: I don't remember his first name, but I remember his last name was Hayes, so he must have been his brother's son or something.

GAM: So, you suggested to Teller that Sid could take over when you left?

TW: I kind of put it this way�I didn't want to say the wrong thing. I think Teller probably asked me if I had any ideas about somebody to do what I was doing, which I reckon was heading the Computation Laboratory. Was that your impression? Who would head the Computation Laboratory if I left?

Teller wanted me to pick my successor or make a recommendation. And I said, "Well, Sid wants the job, apparently." And then, of course, he was a physicist. Now, you know what I'm saying? If you had a degree in physics, you were special at that laboratory. You were the elite. And Sid was a member of the elite. And so, hey, if it's going to grow, if you're going to get more computers, have a guy�you're not going to be happy with somebody�I don't think those physicists out there thought anybody had any sense that wasn't a physicist. Do you know what I'm saying?

GAM: Well, you know, you put the two things together: You said there were very few people who could handle programming and numerical mathematics, and the other problem was that there were a tremendous number of math problems that originated in the physics problem itself. And those guys were busy solving those things, so they naturally took to doing their own mathematics.

TW: Yes, and there are a lot of them that have and did, because there was no other way. Of course, that's why I solved my own problems. I had drawers full of equations. Of course, I've got drawers full of computers, too! Have you seen these things? I mean, they've got the capability of the old Univacs, and some of them are programmable.

GAM: Progress!

TW: But unless you learn to solve your own problems, you're probably not going to find out the answer.

The Univac was started in August of 1947. That was when they started building it�of the first three. And they became operational in March of '51, you see. Now, it was the fall of '52 before they used it on television and all.

When I got to that plant in Philadelphia, I asked them, "Why is that gray curtain all around one of them?" And that was the Census Bureau's machine. I said, "Why is that gray curtain all around it?" And they said, "Well, that's the one we used on TV, and they were screening out all that horribly ugly plant floor up there, so they had that gray curtain backdrop�but it was white when it was shown on the TV. So they had a gray curtain.

GAM: That plant was ugly, wasn't it?

TW: Yes! And it was a bad location, bad plant, bad everything!

GAM: But, actually, a lot of people said that St. Paul Univac was a better place. They had built the kind of computers we wanted.

TW: They had different kinds of engineers and all.

I knew "Pres" Eckert [4]�and I knew him well. I spent the night at his home. And, in fact, when MacArthur was Chairman of the Board of Remington Rand, we went up to have lunch with him one day, and he wasn't there. But we weren't particularly interested in MacArthur as much as we were in talking to Mr. Rand, who at that time was the son of old Jim Rand. Pres Eckert wanted me to tell Rand to back the Philadelphia laboratory and the development of a new line of computers and commercial computers. Meanwhile, St. Paul said, "We are the ones to develop a new line, so we want the millions of dollars it's going to take advanced to us." And so there was competition between Philadelphia and St. Paul.

So, I went up and spent the night at Pres' home. We got up early the next morning, and went on up to the headquarters in Connecticut�Stamford, I think. So, Mr. Rand and I went over in an office by ourselves, without Pres being around, and he asked me, "Tom, what should I do? Philadelphia wants to do this, and St. Paul wants to do this." I said, "Well, what have they told you?" And he said, "Well, I've got this from St. Paul, and I've got this from Philadelphia." I think Philadelphia�I could have them reversed, but one of them was one sheet of specs, like what it would do, and highest speeds and all that, and memory capacities. And the other one was three sheets�probably St. Paul, because I would think they would be the three sheets. They're more engineering-minded.

And I said, "Well, Mr. Rand, you can't take one sheet of paper like this from one plant and three sheets from the other group, and make a decision as to which one to go with! You don't know anything about it. You don't know what you're looking at. You don't understand the problem. You can't make a decision when you don't know what you're looking at. You need to hire somebody that understands these things, that you've got confidence in, and let them evaluate the two."

Little did I know that he was going to take my advice, and go to Bell Laboratories and hire a guy who knew what he was he was doing, and he had confidence in. And he brought him in and damned if he didn't fire half the people in Philadelphia, by a very clever trick. He asked everybody to put in a resume of their experience and background, and when they were hired. Then he went through them. Everybody who had not worked on something like that six months previously he got rid of, because he said, "If you've only been in six months, you haven't learned enough, you don't know enough to really be of help." And he got rid of them. He cut the payroll way down by just firing people that they had hired.

GAM: Did you know Herman Lukoff?

TW: I don't think I've ever met him. The name's familiar.

GAM: He was the guy who made the LARC finally work.

TW: An engineer? That's probably why the name's familiar. Well, you know, Cray and Mullaney and Kisch�all those guys back there�I knew them and had a lot of respect for their capability.

GAM: Yes, they were really great people, and they came through ERA (Engineering Research Associates).

TW: Claire Miller was one who worked with us at Convair.

I arranged to sell the little company to Cuthbert Herd. I was in with a fellow named Dave Willis.

GAM: I know him.

TW: I said, "Dave, if I leave before the end of the year, I'll give you my half of the company." And they didn't really believe I meant it. But I'd already arranged for Cuthbert to buy it. They completed the operation, and wrote me and wanted to know if I would put in writing what my gentleman's agreement with them was. And I did. It didn't cost me but a few hundred thousand dollars, but I felt like, if I'm not going to stay here and make something out of it, I don't want to take my half away from Dave. And I told him, "Well, if I stay for a full year, I'll take my half. If I don't, and leave early, you can have my half. So, my half amounted to several hundred thousand dollars.

It was easy in those days. If you knew the things that Teller thought he would like to teach kids in California�how to solve problems�you could get business. For example, a guy called me one day about a quarter to twelve and said, "If the Polaris missile is sitting in an upright position on the firing stand, getting ready to fire, but during the countdown there's a problem, and we need to check out the wiring to see if something is mis-wired or something, can you do a computer program to know where to probe and so forth? And I said, "Sure. It can be done. Yes, I can do that." And he said, "Well, the programmers here in Lockheed don't think it can be done. Would you come over and talk to them?" I said, "Well, it's time to eat now." He said, "Yes, but come on. I can get them together right now." I said, "Well, I'll miss lunch, but I'll come over."

So, I went to Lockheed in Sunnyvale, and they had maybe twenty or thirty guys in the room. He asked me to go to the blackboard and show them what I had in mind. So, I showed them how to do it, and they said, "Oh, well we didn't know you were going to put it on tape!"�because there wasn't room in the memory to keep all the pin numbers.

But the problem was that they didn't know that I was going to put it on the tape sorted. I was going to sort first, and put it on tape so when I dragged it back in, I was dragging in the right part. And that's the only thing they missed.

When I left there, one of the guys walked up to me and said, "I'd like to come and work for you." And so, I just gave him the contract to do it, because they took it away from Lockheed and gave it to me. So I hired their own guy to do it! He was a capable guy. He understood right away what I was talking about. I doubt if there was anybody in that room that ever sorted, you know. Scientific computation doesn't involve sorting. I sorted the pin numbers.

GAM: Have you kept contact with Jules Mersel at all?

TW: No, not even after I left the Lab. I don't think I ever heard from him again.

GAM: That's the way it is with me, too.

TW: He had ambition, but he wasn't as capable as he needed to be to really get something done.

GAM: I remember you telling me when you went down to Convair, you were telling them how much nicer our offices were at Livermore. And the guy said, "Well, we prefer to put that into salaries."

TW: That's what he said; that was Ben Ferber. And I went out and bought some chlorine disks, you know, at the janitorial service. And I put them in the urinals down there at Convair. It smelled so bad! And it was a dirty, filthy place to work!

Gerkin nearly drowned down there one day. He couldn't swim. We started, I think, at seven o'clock, and we worked, as I recall, a nine-hour day. We'd get off in the afternoon in the summertime, and there was still plenty of daylight to go down to the beach. And so we'd go down there and swim. Jack Rose and I�I loved to body surf, and he loved to swim. So he and I were out there one day, and Gerkin was wading around, knee-deep. But all of a sudden Gerkin was out there where we were in the deep water, and he couldn't swim! So we grabbed him to try to get back to shore, but what had pulled him out there was an undertow.

We were trying to pull him back in, and he was hanging straight down�you know what that's like! He couldn't do anything to help himself�he wouldn't kick right or anything. So, Jack was on one side, and I was on the other side. A little girl about sixteen years old went right by, and she said, "Help me! Help me! I'm drowning!" God, we couldn't even make headway with Gerkin�both of us, holding on and pulling as hard as we could. And she just went by, just p-r-r-r-r�just like that!

So, there was a kid in front of us, a young fellow. And he wasn't but about ten or twelve yards in front of us. The surf was pretty heavy. And I kept hollering, "Hey, buddy! Help us back here! Hey, help us!" He wouldn't even turn around and look! And so I told Jack, "Keep Gerkin up." I figured he could do that. "Just keep him up. I'll swim on by myself and get help." So, I swam as hard as I could, and went I got even with that guy, I said, "You son of a bitch! We're drowning out there�can't you turn around and help us?" He said, "I'm drowning!" So I said, "Excuse me!"

By the time I got to shore, my lungs were�God, I could hardly breathe, I couldn't stand up. But when I got knee-deep, there were some guys there, young kids with surf sleds, and I told them, "Folks are out there drowning�three or four of them!" Boom! They jumped on those sleds and started paddling out there, and they dragged everybody in. Nobody was lost. God, it was close!

I love to body surf. I got a pair of nylon trunks, and I wore them till they were frayed and unraveling. After the Gerkin incident, we decided we'd get in shape. So we stopped eating the planked whitefish every night, and we drank juice for a couple of days, and then took a half a carrot each to work, and that filled us up. And then we'd eat a nectarine every day for lunch. We got in shape, and I believe Jack Rose could have swum to Japan, he got in such good shape! But I'd give out. I can only swim so far before I get tired. But it was more like walking with him�he'd keep right on.

GAM: Well, you ought to learn how to snorkel.

TW: I tried over in Bimini. The water's so clear over there, you can't see it. It looks like you're floating in the air. Jack got in with a bunch that was snorkeling, and a guy went down one day, and he didn't come back up. And another one dove down there to see what had happened, and he said he saw a shark that had him right across the belly. He had the guy in his mouth, and the guy that dove down there couldn't do anything about it. He came back up, and said, "There's a huge shark down there, and he's got the guy in his mouth!" They lost him. And a few days later some joker claimed that he found a watch in a shark's stomach, but he was just playing a joke.

GAM: Gee, that's terrible.

TW: The guy saw it happen, so he knew what was wrong.

We had a couple of guys over from Hughes Aircraft and all, and one said he wanted to form a division in Lockheed to do missile work and get some contracts, you know. And he probably was the one that cranked it up and started it. What he wanted from me was a commitment to deliver, like 1103s or something to them so that, based on having a machine delivery, they could get a contract and would be able to perform because they had a commitment for computers.

And then he said, "Now, if anything goes wrong, I want you to call me. I'm going on a safari in Africa." And he wrote that off as looking for a mountain near the equator from which to launch missiles, but he was just getting his tax deduction, you know? So, the calls were $5 a minute in those days to talk to him. I think I did talk to him once while he was over there.

But Lockheed Missiles and Space Division became a reality, and they wanted to move to the Bay Area. I had helped Thompson-Ramo-Woolridge, which was then just Ramo-Woolridge�Walt Bower-design the buildings that they used. So I went down�you know, you used to be able to do things like this�I just went down and asked them, "Could I have a copy of that blueprint?" And I took the copy of the blueprint up to Lockheed. And the two buildings of Lockheed Research�one of them is a flipped-over mirror image of the other one�are the same buildings that are at Ramo-Woolridge! And they asked me to take them to the architects, which were up on the second floor of a shopping center, near Mountain View. And then I finally wound up working for them at Lockheed Research.

GAM: You know, since you were so active, it must have been�how did they manage to get you to come here and work for the bank?

TW: Well, I was running a little company, and starting it up, and doing the Polaris project and some others. I put out a little newsletter, like a directory of everybody that had a computer and what the equipment was, and the names and telephone numbers, and updated it every month with a little newsletter.

My wife was born in this house, and her grandmother�that's where her grandmother lived over there. Her grandmother died here in this town. That's where the old bank was. When she died, by then I had an unlisted number. You know, there wasn't any way to get any rest if you didn't. People couldn't call because they didn't know how to get in touch with me. I'd written to our folks and said, "We've got an unlisted number, and here it is�you'd better save it."

But when she died, there was nobody who knew exactly how to get in touch with us in California. So, the next day, they had figured it out, and called and woke us up�you know, there was two-hours' difference in time. They called and woke us up, and said, "Your grandmother has passed away, and the funeral is going to be this morning at ten o'clock." Well, that didn't give my wife time to even get back for the funeral. So, she said, "The ten years is up." And it was; we'd been out there about ten years. And she said, "I don't want to be out here when my daddy dies, and I'm not even going to get home for my grandmother's funeral. And I'm not going to be out here when my daddy dies. So let's go home." So, we put the house up for sale.

Now, that was just about the time that I was telling Dave Willis�he was still working for Lockheed Research, and I was starting the company. We'd just started it. And I said, "Well, Dave, I've got to go. So, you'd better stay on here at Lockheed and not even come out." He said, "No, let's just go ahead with it, because you haven't even sold the house yet."

So, that's when I made a deal with him. I said, "Well, look, if I go through a year, stay and not sell the house, then I'll own half of it." "But," I said, "if I leave before then, you can have my interest in it, because I don't feel like I ought to just get in it on this basis."

So then Cuthbert Hurd came along. And, you see, IBM had the old�you got to buy IBM stock for something like 85 percent of market price up to a certain�like a profit-sharing program works today. So, Cuthbert Hurd, having been in there so long with IBM in a good position, and a good man, had a lot of IBM stock. He sold his IBM stock and cashed in for a tremendous amount, and he bought controlling interest in an outfit in New York called Computer Usage. Then he went to the meeting and said, "Meet your new Chairman of the Board. I own the controlling interest in the company. But I don't want to leave Palo Alto," because he loved Palo Alto. And why not? Palo Alto is probably the most desirable town in the United States.

So, then, here's this guy who owns the company in New York and wants to stay in Palo Alto. He came to me and asked me if I would sell this operation to him, because it would give him an excuse, see, to stay in Palo Alto and operate the west operation and let those guys continue with the east operation. So, that's why we sold out to Cuthbert Hurd.

GAM: Did you have much to do with Harold Brown?

TW: I went to his wedding. And then, of course, we had the meetings, like I said, on Monday mornings. He was a very bright guy. He was a little bit, what would you say�he had an attitude of superiority.

GAM: Yes.

TW: He didn't have a chance to mature properly. He got his Ph.D., I imagine, in his early twenties or something, at Columbia or something.

GAM: Pretty early, yes, at Columbia.

TW: And, no doubt about it, he was very, very bright, very intelligent, but he would make snide remarks about other people and things, which�hey�he was right, you know. But, he was kind of smart-alecky, to me.

GAM: Yes. What about John Foster?

TW: He was just straight on, a straight guy. He'd just tell you like it was. There wasn't anything fancy about him.

GAM: I saw him just about a year ago. He hasn't changed�I mean, he looks the same�thin.

TW: He came down on a motorcycle. I think he went to McGill University, didn't he?

GAM: I don't know, but he certainly looked good.

TW: Yes, he headed DDR&E for a while. I had some correspondence with Harold. It was trivial; I kidded him a little bit when he was working for Jimmy Carter as the Secretary of Defense.

GAM: Yes.

TW: That only lasted four years. When he was at DDR&E, when he was Secretary of the Air Force, I called him once. They had put up a missile. They were going to rendezvous up there, and they were going to dock with something that was already in space. And I kind of kept up with it from the radio that day as I was moving around, and I could tell that they weren't going to make it. They were going to run out of fuel before they ever got together.

And so, I called Harold and told him, "Hey, I know what they're doing wrong, and it's going to be too late to do anything about it. But, I want to come back and talk with you about an idea I've got." And he said, "Okay, come on back." So, the next time I flew to Washington, I checked into the Marriott Twin Bridges. A guy from over there called me and said, "Secretary Brown wants me to send a limousine over to pick you up." I said, "No, don't do that. I've been here a hundred times. I know exactly where I'm going, and I'll get over there by myself. Don't worry about coming and getting me."

So, I went over there, and they had a meeting set up. When I walked in, a guy said, "Well, I didn't know who it was from Slocomb, Alabama, who was going to come up here and talk to us." But it was a guy named Miles Standish who had been in the Air Force assigned to the Lockheed project, the missiles, the satellite project. We talked, and he gave me a plot of what they had done. They'd done it strictly Air Force. They just didn't understand the mathematics of how to make the rendezvous successful. They do now, of course. And I'm sure somebody probably told them that night they were having the trouble in Washington. But it's a case of where they don't get down in an organization far enough to ask somebody�

GAM: �who's got to do the work.

TW: Yes, and understands the technicalities of it. And I talked to him about some other ideas. I didn't push him too far. But Miles Standish had been a captain when he was at Lockheed, and that was kind of a�you know, there was a captain Miles Standish in history, and so that kind of made a joke. But now he was a Lt. Colonel back there, so that kind of ruined the joke.

One time I kidded Harold, because they put the heat on the intelligence community at one time, like they're trying to do now, with certain senators claiming that we don't need the CIA and all that. And they had threatened to transfer some intelligence operation to the Department of the Interior, I think, or some screwy thing like that. I think a threat came out last week to do something.

GAM: I don't know, there's a certain kind of creeping death in those places. There are far too many people there now.

TW: I have another anecdote. Teller asked me to help him get a computer for the University of California on some kind of deal from the manufacturer, which would be less expensive�you know, a kind of cheapie. So, I went to�I didn't expect to get it, but I went to St. Paul, and I asked Bill Norris if he would furnish them with an 1103 or something at the university, like on 40 percent of the normal rental. And he said, "Don't get in a pissing contest with IBM! They piss away more than we can!"

So, I went back to San Francisco, where I was working for Remington Rand. And I took a Remington Rand letterhead, and I wrote a letter to Dr. Edward Teller, at Berkeley. And I said in there that Remington Rand will be pleased to furnish the University of California a computer at 40 percent of the normal rental rates, and you'll be able to use it during lunch hour without charge, and we'll do something on the maintenance, and all of that. Pure phony! There wasn't anything to it!

So, I called Teller, and I went to his house�I think, like Saturday afternoon. And in front of his home in Berkeley, I said, "Edward, I can't do a thing about getting you a computer, but read this." He read it. And I said, "Do you know what to do with this?" And he said, "Yes, I understand." You see, what I did was threaten to give the University of California a computer at 40 percent of the base price, which was pure bullshit. He showed it to IBM, and IBM gave him one, and it had to give the whole country one!

GAM: Oh, that's great!

TW: They beat us out!

GAM: Well, yes, I suppose so.

TW: IBM beat us out, do you see what I'm talking about? IBM put one over on us! They got it. We didn't get the computer.

GAM: That's great!

[1] Everyone called it the LMO compiler because that's the way Bob Price pronounced it in his deep southern (Georgian) accent.

[2] Marilyn 1486 was named after Paul Herget's daughter. The Minor Planet Center was established at the Cincinnati Observatory in Ohio in 1947. See also P. Herget, The Computation of Orbits (1948).

[3] Editor's note: That remark, now regarded as one of the stupidest things ever predicted, has been attributed to many different people. It is not clear who said it first.

[4] J. Presper Eckert, with John Mauchly, developed the Univac.