“You have to have a social tolerance for the fact that you are going to talk past each other for a much longer time than is comfortable.  So that’s why you have to put it all on one sheet of paper. You can have all kinds of diagrams, but you need to boil it all down to what fits on a sheet of paper.”           Jay Davis

A Conversation with Jay Davis, Champion of Accelerator Mass Spectrometry and First Director of the Defense Threat Reduction Agency

Jay Davis is a gregarious Texan transplanted to California, and then Washington DC, with his love of homespun wisdom and simple solutions intact. A physically imposing man who looks like he would be much more comfortable tossing telephone poles while wearing a Scottish kilt than fine tuning an accelerator, Jay is always thinking about how to make friends, and how to get along with people. He was the first Director of the Department of Defense’s Defense Threat Reduction Agency (DTRA) and the principal implementor of large-scale accelerator mass spectrometry while at the University of California’s Lawrence Livermore Lab.   

Jay’s experience with custom accelerator design and the physics of proliferation of nuclear weapons landed him a slot in 1991 on the inspection teams after the first Iraq war, looking for the Iraqi nuclear material production facilities. After months of futile inspections seeded by deliberate misdirection of the Iraqi military and weapons scientists, Jay was on the team that recognized the end plates of a large electromagnetic separator being trucked across the desert. The team took chase, and ultimately unveiled the full scope of the Iraq’s attempt to build nuclear weapons. The Iraqi approach used an electromagnetic separation technique abandoned by the US decades before. The Iraqis didn’t need something elegant and highly efficient like the US and USSR to permit them to make thousands of weapons – they just wanted a few, and they didn’t have to be anything but deadly, stripped to the essence of mass destruction. Jay’s ability to see the heart of the matter made this easy for him to identify. He was not looking for the complicated centrifuges of the US enrichment program, but just simple systems that would provide enough separation of uranium isotopes to make a uranium fission weapon. Jay and David Kay’s account of this story in Physics Today is an outstanding example of how to tell a riveting but scientifically and politically accurate science story.

Jay is the president of the Hertz Foundation which provides unique financial and fellowship support to the nation’s most remarkable PhD students in the physical, biological and engineering sciences.  We interviewed him at his office in Livermore.

Why is effective communication difficult for scientists?

I think most scientists forget that when they talk to non-scientists, they first have to explain why they were paid to do this. You need to start your explanation with where it fits in the national portfolio. And when you go to the public as opposed to the practitioners of your art, you have to really start from the beginning, to start all over. We scientists are not good at starting all over.

These days there is a need to be more than your specialty. All the interesting problems are at the boundaries. You have to get a degree in something but the problem you are going to work on is not in your field, it’s at the boundary between your field and another field. Pick any area. There’s something exciting going on there. You have to communicate across boundaries.

That’s why we get the Hertz fellows together at retreats two or three times a year. The guy doing computational genetics listens to the gal doing microfluidics.  I recently visited my old undergrad school, University of Texas, to meet with a group of chemical engineers. I asked what they were doing. One guy was doing what I think of as chemical engineering – electrolytes and batteries. Another was working on nanoparticles for drug delivery and another was working on nanoparticles for structural stuff that strikes me as material science. They are all happily chemical engineers but the problems they are working on are way across the boundaries.

When scientists come out of college they are used to deluging people with how smart they are, and that doesn’t work with most audiences. Too much detail gets in the way of championing ideas. How can they improve?

I continue to learn in that arena. Fundraising for the Hertz Foundation, I’m often in pursuit of the great white whale. I’m looking for $100M somewhere and getting in front of the right people is not easy. Having the right pitch is also not easy. I discovered that I was back in the old lab trap where I was searching for the 20 perfect viewgraphs. I realized that what I really needed to do was put everything on a single sheet of paper. I finally threw out viewgraphs and wrote that sheet of paper, which has worked very well when I send it to people. Every idea that matters is there. And we as scientists have gotten away from that.

Tell us about the history of accelerator mass spectrometry. The concept was started by others, but you built the Center for Accelerator Mass Spectrometry (CAMS) with what differences in mind?

If you had asked a nuclear physicist in the middle 60s if you could count carbon-14 atoms one at a time with an accuracy of 1% relative to a carbon-12 beam, he would have said it was absolutely impossible. It turned out to be simple and it worked the first time. Three groups invented it independently at the same time. There were two papers back-to-back in Science and one in Nature in the same week. All the machines that had done that work were converted from nuclear physics facilities. The machines were compromised by that original intent, and the beam optics weren’t right in many regards. I set out to scratch-build a machine that nobody had ever built. It’s a machine from 1962 except it has an overlay of 1980 electronics on it and it has beam optics unlike any accelerator in that class. It was a spectrometer, not an accelerator. As one of my Australian collaborators once said it was the first machine built to do something more than simply hose the target.

My most useful scientific insight was pushing this pure physics technique, accelerator mass spectrometry, into biological and clinical medicine. I just pushed it across the boundary. We had hilarious meetings where you’d get a nuclear physicist and nuclear chemist together with a toxicologist and an MD and you’d try to design experiments. Whatever you thought going in was not at all the experiment you got coming out. It was very humbling.

You have to have a social tolerance for the fact that you are going to talk past each other for a much longer time than is comfortable.  So that’s why you have to put it all on one sheet of paper. You can have all kinds of diagrams, but you need to boil it all down to what fits on a sheet of paper.

How did you champion the idea that the accelerator mass spectrometry could change medicine?

First, it was something that could only be done at a national lab. There were absolutely no rewards for doing it at a university. The physics department would not be interested in medicine. The medical department would have said, “Holy cow, an accelerator!” The first machine we built filled 7,000 square feet. The timing was unique. It was just when Lawrence Livermore Lab started an internal, long-range funding program. The original criteria for this laboratory directed research and development was to cross disciplinary lines.

I had a good pitch. I said, “90% of life is carbon and the next 5% is hydrogen or tritium H-3, hydrogen with two neutrons. I can measure carbon-14 and tritium at sensitivities a factor of a million higher than any technique you have.” You don’t need to know a lot more than that. That’s enough to investigate this topic as having extraordinary promise. I can take an isotope you routinely use in your laboratory and find it a factor of a million more sensitively, and by the way, that means I can freely dose human beings with it at levels we can easily measure. I can freely do human experiments. If I give you carbon-14 tagged aspirin, you are in more danger from the aspirin itself than you are from the radioactive carbon-14. That was a pretty powerful argument.

The way you speak about carbon-14 and aspirin is something we keep hearing in stories of good science communication. There is some easy-to-grasp concept that goes with telling a compelling science story, and creates a context for understanding.

Give people something that they can immediately wrap their heads around. You need to give people something that makes sense and fits in their worldview.

You then had a tool and you had to convince people like the National Institutes of Health that it was worth moving forward. Tell us about that phase.

That was tricky and it had a funny backspin. In the lab we demonstrated the technique could work. Like most scientists, we thought the biomedical research community would just fall all over themselves to get at this tool. We thought it would be the most wonderful thing in the world because we could effectively give it away free for a while due to our developmental funding. But it turned out that the barrier with biological research people was very high. They already were staffed with scientists and were already funded by National Institutes of Health.  And their answer was, “Oh god, we’ll have to get new post-docs and learn some new lab techniques and write a grant that’s different from the one we have been getting renewed regularly”.

But the people that were receptive were the medical practitioners who were frustrated with the tools they had available. They were engaged with sick people and they were pissed off that they didn’t have diagnostic tools. They turned out to be the ones who grabbed what we offered by the scruff of the neck. They pulled it forward to the point that the National Institutes of Health would finally accept a proposal, and so help me god, they funded an accelerator inside a weapons lab for biomedical research!

This is a great story, because on November 12th 2013, CAMS defended the fourth cycle of their NIH proposal for about $8M over a five-year period. And the NIH is funding two more small accelerators to go into that lab. They will have had the first, second, and third biomed machines – primarily because of pull from the clinical community.

That experience underscores the importance of enlisting support.

We were very careful to avoid the typical mistake of saying, “We want to own it because we came up with the idea”. Our attitude was, “How fast can we push this out of the lab? Because our post-docs can follow it out.” There are ten or 12 accelerator mass spectrometry labs around the world now staffed with people who were a post-doc with me or worked for the lab I started. We sent them all over the world. You also want industry to take this stuff. Three companies were started on our patents.

I feel great about it. Not many men and women get to create a field. I used to say that my career was based on helping people who could do things I didn’t know how to do, do things they didn’t know they could do. I couldn’t be happier about how this came out because it was much bigger than I ever thought it would be.

Talk about collaborators a bit.

It comes back to this question of ownership. Do you want to do something you totally own yourself and control? Or are you interested in working at the margins in such a way that you and somebody else can share a problem and do something that neither one of you can do alone? The people that are interesting fall in the latter category.

The business I was in was competitive and inherently collaborative. People worked in each other’s labs and copied each other’s hardware. We learned from each other how to do things better. Nothing was locked up and hidden.

We were lucky with CAMS because the Regents of the University of California invested a quarter of a million dollars in it in the early days. I talked the Livermore Laboratory into signing a memorandum of agreement that said, in exchange for that investment, the faculty and staff of the University were guaranteed access to the facility. So if someone said, “Why is there someone from China in your weapons lab? And why is there someone there at midnight?” we had it covered. There is a piece of paper you signed that said they needed to be a student in good standing. We had a big leg up on getting collaborators from all over.

From the medical perspective, you have got to have MDs at the lab to touch a human body. So from day one, if you wanted to work in that world you had to be a good collaborator, because our lab had no physicians. The hard part was to figure out what your collaborators were capable of doing in the lab. More than half the people weren’t mentally rigorous enough in the physics and experimental aspects to master the hurdles. Partly it was because they just couldn’t believe the sensitivity of the technique and they couldn’t cleanly handle low-level amounts of radioactivity. They would flunk and fall away. It’s a pretty damned Darwinian system. The third time you cross contaminate your samples, I am not willing to carry you. All the art is in getting the sample and preparing it. You have a harder time finding good collaborators than you might think.

Let’s talk about your experience at DTRA. You certainly would have had opportunities to deal with new ideas there. What worked well?

The Defense Threat Reduction Agency is not a research agency. So when it does R&D, it is a little bitty R and a big D. I would get up every morning and stand in the shower and say, “I am not running a national lab”. The people I had were superb executors but they were not the creators of the ideas. What would happen is that the contractors who supported my folks would come up with the ideas. Then my folks would come and try to sell those ideas to me.

At DTRA, there was a lot of relationship building. You learn that you’ve got to be on message and absolutely say the same thing all the time. If you try to send a different message to a different constituency, those lines will close behind you so fast and your credibility goes to zero. I was better off to look at somebody and say, “You’re not going to like what I’m going to say but this is my position on this”, than to try to tell them something that makes them feel good, and have them discover that I said something very different two offices down. And you have to be infinitely sensitive to the fact that anything you write or anything you say can instantly be on the front page of the Washington Post the next day.

I’ve had a ton of media training. The first thing you learn in media training is for god’s sake, talk in sound bites. Turns out, that’s all the training you need to work with interpreters when you are overseas, as I often was with DTRA. You kill your interpreter if you talk too much.

What’s your advice with DOD? How do scientists think about that space better?

Look for opportunities to contribute. Take on additional duties other than what was assigned. My approach was to completely deinstitutionalize it, to go to the labs at large and say, I want the ability to pull 100 of your best people for a week and I’ll pay for it all.

You build personal relationships that have long-time constants. If any of this stuff is short term you’re dead.  All these problems are based on long-term relationships. Think of it as a tenure decision. You can have a six year run on things that matter.

You were the front man for a big organization at DTRA. How do you deal with acknowledging the work of all those people who supported you?

I’m so old, I have not had my name on a refereed professional paper in over 20 years. During that time I have probably given 5,000 viewgraph talks.  That said, if you did not put my name on the publication I would be outraged, but on the viewgraph talk I’d be fine with being verbally mentioned as a collaborator. That’s probably good enough. Nobody should list all their collaborators at the front of a viewgraph talk. It’s understood you are giving a summary.

I guess that is the distinction. Be enormously careful with what goes into archival literature and more forgiving in what goes up front on a set of viewgraphs. In my world the assumption is you are just representing the work of other people. For three years I ran a 2,000-person defense agency and I never gave a talk on my original work, and that was understood by everyone in the agency.

Tell us about ways to talk to Congress.

First of all, you do most of your talking to staff. When I was heading DTRA I had the status of a Lieutenant General, which is a three-star General. You’d think that was something, but political appointees have the status of four-star General. Since I was not a presidential appointee, with rare exceptions, I would not get a face-to-face meeting with a member of Congress. You are not worth beating up, and you do not speak for the Administration. You do not speak about policy, and policy is 90% of what Congress wants to talk about.

Your influence is mostly through the congressional staff by speaking to them about a problem they have today. They are not looking for you to bring your portfolio in. The first rule when you go up on Capitol Hill is that you have one subject to talk about. You are asking for one thing or helping them with one problem.

It’s not a scientific engagement. You’re not going to win with equations. It’s a place where you really learn to slow down, to slow your science and communication reflexes down. You’re not allowed to be insulting.  The first time I went to testify before Congress I was told there are two things to remember. First, your chair is lower than theirs. Second, if the Senator from Connecticut wishes to throw a tire and have it bounce off you and hit the Senator from Ohio, that’s just fine but you don’t get to throw a tire.

It’s an interesting situation. The Pentagon has a five-year planning process. So if you have some wonderful new idea, you are dealing with the reality that the immediate budget is fixed. It cuts both ways. One is you have to give your boss a superb argument for why you need more money. And since I had a $2B budget I would never go to my boss and ask for $5M. “That’s 0.025 % of what you have, get out of my office”. Not a good approach. Whereas if I had a $50 or a $100M idea, I better have really good leverage because he or she is going to take that money from someone else.

I could never get the labs to understand, “I can’t get you money today. Tell me what you are doing and I can give you $25M in two years.  But I need two years to get there from here. If you can’t be patient and you don’t understand there’s a kind of a long run up, I can’t help you.”

When you think about putting a presentation together, what’s your approach?

I am always careful to try to avoid monotony. My favorite trick is to put words on one viewgraph and a figure on the next so the viewgraphs are different and there is a little bit of awakening each time. You really have to simplify, simplify, simplify. If you get more than about 60 words on a viewgraph, you’ve killed your audience.

When I first retired, for several years I ran the Valley Study Group, which is the group of retired engineers and scientists from Lawrence Livermore and Sandia labs. They would meet once a month for dinner and a talk from an outside speaker. I learned very quickly that the lab people would bring in a talk that worked well in a small conference room. They unconsciously set it to give to 15 to 20 people around a table. But when you get in a room with 180 people, your lab talk dies because the dimensions are different. Your print is too small and you have too much text. I would have to explain to speakers that you have to throw out your talk and remake it for that room, because most of us don’t give talks in venues like this. And these people don’t hear very well and their eyes aren’t all that good either!

How do you structure your presentations?

I worked for a retired admiral who might tear the title page off and take the first or the last viewgraph off and hand it to me and say “that’s your talk.” So for him the first and last viewgraphs were the same, and had to have everything on it.

Physicists have the problem of what I call deriving the answer. They want to walk you through the derivation. I’ve found it is much better to assert what is important and then defend it. Give me the answer and the assertion up front. “I’m going to tell you I can do the following, and it will cost you this much money, and I can do it in that many months.” It’s fair to challenge me and undress me to make me defend my assertions, but it is an insult to you to have to sit there for an hour to see that in the last viewgraph.

It’s funny, I have multiple hats now. I’m just finishing my year as chairman of the board of directors for the Chamber of Commerce in Livermore. We did a survey of skills that companies around here need. We divided them into primary and secondary skills. There are a lot of manufacturing companies around here.  The secondary skills were, “Can you speak, communicate and write?”. One employer came back with: “Would I feel comfortable with this person alone with a customer? We can teach the primary skills. I can teach a kid to do CADCAM.  I can teach statistical analysis but I am not going to teach language and communication”.

Who are the great communicators, the speakers who moved you?

I have a funny list.  One on the list is a set of committees.  There is a wonderful book called God’s Secretaries, The Making of the King James Bible, published a couple of decades ago.  It describes how James and his advisors set up a system of broadly representative committees to translate from whatever original or near original sources were available and argue the proper interpretation in Jacobean English.  Another committee than reviewed the work of those committees to produce a final authorized translation.  Most of us would assume that was a mechanism to assure mediocrity, but it crafted a marvelous style that set the standard for our language – and let James accomplish a very sensitive political end, reconciling the Calvinist Scots Kirk with the much more Latinate Church of England.

As to real people, my heroes are Shakespeare, Lincoln and Churchill. Shakespeare because his prose is written to be spoken, not read, and because he invented hundreds if not thousands of words that we truly need. Until you heard “incarnadined” in Macbeth, you didn’t know that you needed that word.  Lincoln because more than anyone ever he managed to speak for a Nation in times of trouble. Even his humor, with its dark edges, seemed to have a teaching purpose beyond what others manage.  As a teenager, any time I had to give a speech, I’d stand in front of a mirror and run through the Gettysburg Address several times just to try to get and hold a meter that sounded good. The only one I know about in a quantitative sense is Churchill. If you look at his magnificent wartime speeches and you run a computer assessment of the vocabulary, the language compresses way down to simple basic Anglo-Saxon words, everything new or anything adopted from other languages is dropped out. The emotionally loaded words tend to be the simplest, oldest words in the language, and he knew how to play them like an organ.