Meet Konrad Hinsen
Research Scientist at the Centre de Biophysique Moléculaire
Associated Scientist at the Synchrotron SOLEIL
- In what “slice” of CiSE do you work?
At CiSE, I take care of the Scientific Programming department, together with Matt Turk. My research work is about molecular biophysics, in particular proteins, with an emphasis on method development.
- What sorts of changes have you seen in the field over the years?
I have witnessed the enormous growth of computing everywhere over the last 30 years: in science, at home, in business. Scientists got ever more computing power, but with the move from mainframes to commodity hardware they lost the help of the specialists at the computing centres. We had to become our own system administrators and programmers. Next was the transition to parallel computing, which lead to a partial transfer of work back to computing centres.
Concerning software, the possibility to use scripting languages and graphical user interfaces has led to a democratization of computing in science that had its good and bad sides: easier access to advanced techniques, but also increased use of computing techniques by people who don’t fully understand their limits.
Finally, concerning the scientific practice in general, computing started out as a limited tool for well-defined subtasks (computing an integral, solving a simple differential equation), but then grew in importance to the point that today we have entire domains of research based on nothing but data and algorithms. Such research can easily get out of control, because it is very difficult to verify the results of enormous computations. Scientific validation requires comparing to observations and experiments in the real world, but today we compute many things that are not even accessible to experiment. Is that still science?
- What is the most exciting aspect about your work for the near future? The far future?
In computational biophysics, virtual experiments are coming within reach. A virtual experiment is a simulation realistic enough to compute exactly what is being measured in a real experiment, including characteristics of the instrument and unwanted features of the sample. I hope that this will boost both simulations, which will be better validated and thus more reliable, and experiments, which can be better prepared through simulation.
In computational science in general, I am excited about the new technologies for collaboration, and for sharing and publishing computational methods and results in machine-readable form. In the long run this will lead to better science, but this also requires changes in the attitude of the scientific community.
- If you were to explain Computing in Science and Engineering (either the magazine or the field(s) it represents) to a five-year-old, what would you say?
It’s video games for grown-ups, who use them to explore the world through an enormous microscope, or inversely the world seen from a large distance.
- Big Data… What’s more exciting or important (or is there anything more important)?
What’s important is doing good science. I think it’s too early to say if and how Big Data will help with that. Big Data techniques have led to some spectacular success stories in solving specific problems (Google services such as translation are perhaps the best-known applications) and it seems quite probable that they will revolutionize some fields of research, but we will have to be patient for a few more years before we know. In the meantime, these techniques are definitely exciting.
- What is one thing that would fundamentally change the average person’s reality if he or she worked with you day to day and saw what you saw?
When talking to people who have no contact to research, I often notice that they have an idealized view of science, which they see as a source of ever increasing certain knowledge about the world. If they worked with me day to day, they’d see that scientific discovery is an erratic path along which every answer found raises a new question. They would probably have both more and less confidence in science from then on. More because it’s demystified, less because they realize it’s not perfect.
- What is the most important application of HPC/computational science/data visualization in your opinion? (Protein simulation, climate modeling, etc.) Why?
I wouldn’t single out any one application as the most important. Protein simulations and climate modeling are important, but so are biological simulations on a larger scale (protein interaction networks, cells, cell colonies, …), simulations of new materials, or the analysis of social interactions.
- Conversely, what is the scariest?
Again I have more than one candidate. Massive surveillance, as practiced by the NSA, is one of them, but massive analysis of data published voluntarily on social networks could be just as scary – imagine what your bank or your health insurance could try to deduce from that. Personalized medicine could become scary as well – do you really want a computer to tell that you have a 40% chance of dying from a heart attack in ten years?
- Why do you do what you do?
I like the intellectual challenge in both science and computing. And the idea of contributing, if only a bit, to a better future for everyone on this planet.