This talk was part of a panel entitled “Nerd talk, geek speak, and the challenges of 21st century knowledge silos” held at Arizona State University on October 1, 2012, and hosted by the Center for Science and the Imagination. Ed Finn is the director of the Center for Science and the Imagination and chaired the panel. Neal Stephenson, well-known author, was also on the panel.
Introduction
When Ed first called me a geek I was insulted. Then he said Neal was a geek, too, and I felt a little better. As a product of the 1950s who was packaged and labeled in the 1960s and 70s, I generally associate the words “geek” and “nerd” as negative and I have to confess that I spent a fair amount of effort in my early years trying to avoid being or become one (or the other).
Failing at that, I signed up for engineering school in college and learned a bunch of stuff that seemed perfectly normal, among others who shared that view. To put a sort of exclamation point on it I went to graduate school and earned a Ph.D. in engineering. I went to Berkeley and everyone seemed pretty normal to me there and the things we learned seemed pretty normal, too. I will admit that the further I got, the more delightful and exciting the ideas became.
After graduate school I launched my career as a professor of engineering and I spend my days thinking what I consider completely normal thoughts and conveying the delightful ideas that I have learned to young people, most of whom seem normal to me.
About twenty years ago I met the woman who would later become my wife and I remember one conversation where I explained my research in computational mechanics and structural engineering to her. It was a pretty elaborate description. At the end of making my “case” she looked at me and asked “Why would anyone want to do that?” We were married six months later.
I did take the time to Google the terms “nerd” and “geek” to get ready for this event. Once you get past circus performers who bite the heads off of chickens and snakes, you get to a pretty sensible distinction. Geeks have laser-like focus on something, but not necessarily high intelligence. Nerds have high intelligence, but not necessarily laser-like focus in their interests. For people who do math, science, and engineering—subjects that most in our society view has “hard”—there is often little distinction made between nerd and geek.
So, I am a geek, for sure. If knowing a little math, science, and engineering is an automatic pass to high intelligence then, “thanks” I guess. But that seems pretty stupid to me.
I did eventually figure out that the stuff I know about is not that normal and the way I talk about it is not actually accessible to most people. I also realize that, as a professor of engineering, I have to take some responsibility not just for living in a silo, but also for indoctrinating the next generation. I actually make my living luring people into the engineering abyss. But in my own defense, they seem to go willingly.
Knowledge silos are part and parcel of geekiness and nerdiness. I have observed that there is a clear upside and a clear downside to the silos that I have experienced in my career. Let me just call them “good silo” and “bad silo.” The good silo is where experts go to talk efficiently about complex things. The bad silo is more of a “tick tock the game is locked” sort of mentality that responds to our more primitive tribal instincts. When you get right down to it, I came to ASU because I was having second thoughts about the silo thing and it seemed like this guy Michael Crow was really working on it.
Rocket Center Geeks
I had spent the ten years prior to coming here working on a project we called the Center for Simulation of Advanced Rockets. Our goal was to do a computer simulation of the first 126 seconds of a launch of a space shuttle. That is the period of time that the solid fuel rocket boosters are firing. For those who recall the space shuttle, those are the two smaller cylinders on the shuttle. It might seem like a modest task, but that is an extremely complicated multi-physics simulation problem and our first estimate on how long that computation would take with the best algorithms available on the fastest computer on the planet at the time was 13 years.
This project was funded by the Department of Energy and it had a social experiment component to it—could researchers from different disciplines collaborate on a very large project like writing the computer code to do such a simulation? The reason that they were interested was because the National Laboratories had been largely unsuccessful at getting researchers to talk through the walls of their silos.
But let’s get some perspective here. We were seeing if structural engineers could talk to and work with mechanical engineers and computer scientists. But the problem operated at the level of all that was known at the time. It was a hard problem.
Well, we got to working on the problem and it soon became evident that the Eulerian codes that the fluids geeks used were not compatible with the Lagrangian codes favored by the solids geeks. We quickly landed upon the idea of using staggered schemes but found that most of them were unconditionally unstable. The solids people discretized their differential equations with finite element methods and the fluids people preferred finite volumes with large eddy simulation methods. The numerical schemes blunted the sharp shock fronts, which was clearly in violation of Huygens’s principle. The computer science geeks were concerned with leakage and wanted to do parallel I/O because they were concerned about the volume of information that the teraflop computations would generate.
Okay, so maybe you didn’t get all of that, but suffice it to say that each geek’s concern was very important and very real and it takes a lot of energy to explain those concerns to someone who has not been immersed in the problem for ten years. There really is a part of the silo thing that is hard to get around. When you are working at the frontier of knowledge there just aren’t that many people who can make the journey with you. And when you are down deep you need to be able talk efficiently. The lesson from the rocket center was that it was surprisingly difficult to communicate across silos at that depth because the languages were remarkably different. We succeeded mainly because everyone was focused on a single tangible application—the space shuttle. It provided a reason and a vehicle to operate across silos.
When the money stopped flowing, we all went back to our own silos. But I don’t think any of us forgot what we learned.
Structural Engineers of the Future
I have recently found myself on a committee tasked with imagining the future of structural engineering (or so they told me to lure me onto the committee). The world of real structural engineering is a different kind of silo and I sort of go there with a guest pass because I am not really a structural engineer, I am a professor.
Our conversations in this group are not the deep technical stuff that is facilitated by the good silo. Rather it is about the future drift of the profession as it tosses about on the sea of change. It is a conversation about really broad, common sorts of stuff—things that are true of all professions. But it is clearly a conversation taking place in a silo. And the most interesting thing about these conversations is that they are mostly about building the silo higher and narrower. They are about strategies to exclude others in service to the idea that to do so might increase the importance of structural engineers and thereby the prestige of the profession.
This is the bad side of silos. There is an element of tribalism in the silo mentality. We circle up for solidarity and for social legitimacy. Many of the people in the silo really don’t want to do the deep dive, but they also do not want to get out. The silo protects them from all of those things they don’t know. And as long as you can make a comfortable living not knowing those things, then why not?
Academic Institutions Land of the Nerd, Home of the Geek
I have also spent my life in the academy. Most academic institutions have made geekiness into a hard requirement of its members. And, furthermore, we punish violators without mercy. Those people who get tenure are generally the most focused—i.e., the most geeky. The easiest route to success in the academy has long been relentless pursuit of a narrow discipline. Almost everyone acknowledges that crossing disciplines is treacherous territory for young academics. Some will tell you that it may cost you your career. That single feature promotes academic silos that fail to foster a cross disciplinary, integrative, collaborative environment. And the world sort of needs to have that collaboration.
The great thing about ASU is that it is working on this problem.
The Tower
So a year ago I am just another geek minding my own business when Michael Crow forwards me an email from Neal Stephenson. Neal is looking for someone who might help with this crazy idea of building a tower 15 to 20 kilometers high to launch rockets from. I sent Neal a “how can I be of service” sort of email and he sent me a couple of papers describing the possibility of constructing a tower of this height.
Remember, I am a structural engineer. In my professional silo what we know is that the tallest structure built by humans is the Burj Khalifa in Dubai which stands 828 meters high. And we know that is a significant jump from the previous tallest building Taipei 101, which stands 508 meters. From a structural engineering point of view there is plenty of reason to doubt that a 20 km tower is even possible.
But also remember that I am not a real structural engineer, I am an academic. So I get curious and start doing some computations to try to decide if this tower is even possible. I have been sucked in pretty far on this one and right now I may be the only person on the planet who has done these calculations. And, I am pretty sure I would never have done them if Neal would not have asked. It now seems sort of odd that after nearly forty years in this business I had never thought to ponder the question of how high we could go with a tower.
I would have to say that I have not really left my silo for this project. It is more that Neal has wandered into my silo, pulled up a chair, and sat down for a conversation. I am pretty much at home. I do have the challenge of describing the outcome of my calculations in ways that are intelligible. And so I am kind of always on the alert to that fine line between “silo-speak” and language generally recognizable to non-engineers. The imperative to avoid silo-speak brings me a fresh perspective on my silo—a perspective I hope will improve my teaching and a perspective I hope to share with the professional structural engineering colleagues as they chart their course to the future. The efficiency of being deep in the silo is important, but it is also important to come up for air. Communicating differently about the ideas has caused me to think differently about those ideas.
There has been an unexpected positive outcome to the Tower project for me. I have spent my life biting my tongue. When my enthusiasm is just about bubbling over on an idea I am working on I have never found anyone outside of my field to be even remotely interested. The perspective of my future wife—why would anyone want to do that?—has held sway for most of my geeky life. So I try not to talk about it.
But I have talked about the Tower to quite a few people by now and, without exception, every person I have engaged on the topic has immediately gotten interested in it. Everybody thinks it is cool. People just start brainstorming on the concept without fear and without technical background. I am convinced that this is the result of how big the thought is.
So good silos are where geeks and nerds go to solve hard problems. Good stuff—but hard to get out. Bad silos are where the unimaginative go to hide. Bad stuff—also hard to get out. I wonder if we might be able to unlock the silos with big enough ideas.
Keith Hjelmstad is Professor of Computational Mechanics in the School of Sustainable Engineering and the Built Environment (SSEBE) in the Ira A. Fulton Schools of Engineering at Arizona State University.