Tell us a bit about yourself. What are you most passionate about? Hmm! I'm probably most passionate about enabling helpers and joy. This comes from a perspective of mortality and pessimism: I've got probably under 50 years of being able to do stuff, and there are lots of problems locally and globally that I feel powerless to fix. One tangible facet of being an educator, however, is empowering people to solve problems and teach, too, and that makes me hopeful some big problems get solved. It's also easy to forget that hilarious things are ubiquitous, and there's joy to be found in finding them, in discovering things, in helping others in spite of things not being perfect. I think it's very important to experience joy, especially in ostensibly "professional" settings, in order to do sustained work on hard problems. One of your research goals is to “leverage thermodynamics for societal good”. What are some examples of this? The sorts of things my team is trying to do include projects like discovering chemicals that are good ingredients for inexpensive solar panels, and understanding how the polymers used to make aircraft composites can be used to make stronger, lighter, more efficient aircraft at lower cost. In both cases, the properties of the materials, whether they're strong enough for airplanes or good at converting light into electricity, depends on how the molecules making the materials are arranged. The laws of thermodynamics govern how those molecules arrange themselves when we mix them together, and so we work hard to understand how to exploit those laws to work for us when making new materials. But the ultimate goal of the new materials we aim to invent is for them to be easy to make without harming the planet, reliable, and useful in helping people live safely and sustainably. Experimentalists typically run experiments in labs to test how materials work. Theoreticians do calculations to predict how materials work. How would you describe how computational scientists work? Is it more experimental or theoretical? There's wide variance in whether computational scientists lean more experimental and theoretical, and it's possible to do lots of both! Most of the time on my team, the work we do is more experimental: We set up an experiment using a computer, perform the experiment, and have a look at the results in the same way that someone using a fume hood or glove box might. We can do more experiments in a shorter amount of time, often, but when we're not working with theory to develop new kinds of experiments, we're mostly performing experiments using computers as the apparatus. High-performance computing using graphics cards dramatically expanded they types of simulations scientists can do. With the rise of machine learning and the potential coming of quantum computing, what do you predict materials research will look like in 10 to 20 years? There's potential, and I am hopeful, that many arduous, important, but uncreative tasks will have been replaced with fast computations. For example, the creation and evaluation of the rules by which atoms interact in simulations is something we set up by hand right now, is very fiddly, and there's no fundamental reason we couldn't train a machine to do this for us. It looks like there's great potential for quantum computers to assist with the prediction of the electronic structure of molecules, which would put a whole set of techniques out of business. But I don't have a good sense for, theoretically, whether some fundamental problems in materials science are as "easy" as playing go for teachable machines. In 2015, I thought computers beating top professional players at go was 10 years away, and it happened in 2016. The successes of AlphaFold are super inspiring and I'm hopeful to see something similar be usable in predicting the structure of, say, polymer materials, for example. In summary, I think that in the next two decades it should be a lot easier to to predict the structure of materials from their ingredients, and because of this, those that can predict properties from structures or those that can figure out how to synthesize those structures will be even more sought after. The description in your Twitter account lists bikes and Go as interests in addition to science. How important is it to have interests outside science? Do these interests ever overlap with your research? Without some balance it's possible to ruin anything. Like, eating one slice of pizza is great, but having to eat 500 in one sitting is terrible. The research I do gives me time away from riding bikes, which gives me time away from playing go. I love doing all of those things, but if could only do one of them I expect I'd have under 20 years of attention before I burned out at it. So they're different enough that I can get a break from one when doing another, but there's a ton of overlap between them! Fixing bikes and debugging code are identical processes on different machines. Finding the best move on a go board and finding the energy-minimizing structure of a polymer are essentially equivalent mathematically. Winning a bike race and go tournament both require training, strategy, and split-second decision-making. In all three cases you can be a professional, and my experiences interacting with experts in each domain inform how I think about teaching, mentoring, competition, and supporting communities in the others. How would students describe what it’s like to work on your research team? You should ask the students on our team! I hope they'd say it's a fun, supportive, interesting place to work and learn. If you had to give a high school student interested in STEM careers advice, what would you say? There are a ton of different careers that scientists, engineers, and mathematicians can do, and many of them you'll find interesting and fulfilling. Find them and do them, we need you! It's also a sad truth that many of the systems and practices in STEM education make students feel excluded from their classes, departments, and universities. Universities aren't immune to sexism or racism, and there's a lot of work to be done to make STEM opportunities more equitable. If you feel like sexism or racism hasn't mattered to your education so far, I'd invite you to use that privilege to help effect change. Reading broadly around the history of science and education is just as important as reading broadly about the the topics that got you interested in STEM careers. If you had to hire a nanobot to do a job for you, what would the job be? Delete emails before they reach my inbox? Haha, no, maybe turn things off that are wasting energy? OK. So, nanobot, huh? I'm interpreting "nano" as some things whose size is less than say 500nm, and "bot" to mean things that are programmable. So this is like 10x to 100x smaller than a red blood cell, so it'd be great to have some in my arteries disposing of plaque, allowing me to eat as much jamon iberico I want. But if we can program trillions of them to build more of each other, who can place the molecules in the right places for custom materials, then we're done, right? This is a tall order, though, because I know how hard it is to program things correctly, and how fiddly things are at nanometer scales, so I worry that half-reasoned attempts to do this end up causing more havoc than good. Eric, thanks so much for taking time to talk with us! I look forward to reading your future publications! Want to keep up to date on the latest happenings at DNP123? Subscribe to our newsletter.
Comments are closed.
|