Making Trouble — Model It, or Make It Modelable

This summer I had two great interns in my lab. As usual, they taught me more than I taught them. One in particular helped me refine my thoughts on the theme that pervades my every day.

When Geoff first arrived in my office, I described his summer project to him. I had written 2,000 lines of code in MATLAB to model the Ackerman steering geometry of a tilting (leaning) vehicle. He was to take that code, check it, improve it, and finish it, and we’d build the vehicle as a tilting, steering, cargo-carrying tricycle. He said, “It’ll take three days.” I countered, “I’ll bet it takes six weeks.”

Geoff dove into the code. He only looked up from his computer for two reasons: to hear instructions for using the vintage hand-pulled espresso machine, and to go stare at the physical prototype of the tilting trike to orient himself to the problem. He missed his own three-day target, but crushed my six-week estimate when he proudly showed me the first working computer model in just two weeks.

And that’s the theme pervading my whole life right now: computational modeling.

Why? Here’s how I see it: Galileo Galilei arguably did more to usher in the scientific revolution than any other. The quote “Measure what is measurable, and make measurable what is not so” is attributed to him. In my mind, 19th- and 20th-century science did exactly that, and the scientific method — the cornerstone of thoughtful progress in knowledge — is heavily dependent on good measurement.

We measured everything we could about the scale of the universe. We probed the atomic then subatomic structure of the elements with incredibly elegant, single-parameter experiments that isolated things like the mass of an electron.

This type of science has been so successful that now it seems the true frontiers of science exist less in the study of easily reducible, measurable things, and more in the study of complex, multiparameter systems — like biology, climate, metabolism, and ecology. In these systems, understanding is built with models that can be tested for their validity and correspondence to the messy, complex real world.

Like many physical systems, there’s no perfect answer to a tilting tricycle, only “optimal.” You can optimize the parameters, but because of the limits of physically realizable machines, it can only ever be “almost perfect.” Success is being closer to almost perfect than other people’s models.

But here’s the beautiful thing about modeling. Computational models are digital, and that makes them inherently shareable, independently verifiable, and easy to collaborate on and improve.

Whereas my inclination was to immediately start to build something physical, Geoff’s approach — the approach of a new generation of engineers and scientists — was to begin with a model. Start with bits. Make them perfect, beautiful, defendable, sharable bits, then render them physical once you’ve reached an optimum. Sure, someone might figure out a better optimum one day, but because they can start with working, executable code, they’ll get to it faster.

There’s an even more important reason to encourage this culture of shared models. The more people who have experience simulating the world with success, and making things from those models, the more people will trust in the models of our physical world that will guide how we shape our future.

I read once about the science of perception and the humble practice of catching a falling ball. A ball moving 60mph travels almost 90 feet in a second. The only reason we can catch it is because we have a mental model of where the ball will be when our hand intercepts it. Throughout the course of our lives, we’ve built a mental computational model, which we’ve refined thousands of times, that helps us predict the future position of a ball so we might catch it with our relatively slow reflexes.

We have enormous faith in the ability of a professional baseball player to model the future of a ball, under complex windy, rainy, and noisy conditions, and to catch it. It would be nice to build similar public faith in the ability of our professional scientists to model the future — the future of the oceans if we continue to pollute them with toxins, of the atmosphere if we continue to emit carbon dioxide, and of other problems that require humanity to have a faster response time than its cultural reflexes.