Engineers who are citizens

After a 2-month hiatus, yesterday's talk by David Douglas inspired me to write again. The talk was entitled "Citizen Engineers, Computing and Clouds". So I went expecting a talk on regular folk who do engineering, perhaps the open source community, perhaps rapid prototyping in your home, perhaps harnessing the ideas and contributions of hobbyists for a greater good. The abstract mentioned engineering and computing's relationship to current social and world problems, specifically environmental impact, so I went hoping for a hint of an idea as to how computing, computer science, could help or hurt sustainability and the fight against climate change.

It was an interesting talk, but Douglas's "citizen engineers" are not regular folk doing engineering in their free time (as one might have guessed by extension from "citizen journalists"). Instead, I heard some interesting stuff about the carbon footprint of data centers (on a par with air travel), the lack of openness about power consumption and carbon footprint by Google, and how companies should incorporate environmental impact reduction into their business practices lest they go out of business.

Which is all very well, but here is what I would like to know. Are there advances in computer science that can be immediately applied to sustainability problems, and how? Is cloud computing better for the environment than older forms of computing, and how? Some obvious things were mentioned: data center computing doesn't require nearly as much plastic components as end-user computing, because who cares what the machines look like inside a data center? But here is what I would have liked to hear and didn't: here's a sustainability problem -- you can cast it in this very cool new way which immediately suggests a computational solution.

Maybe the deeper wisdom of tackling world problems with computing is that you can't address any abstract, computational aspect of any of them without thinking at the same time about very physical and mundane things like amount of plastic and power required. I'm thinking of this as the analog to the embodied intelligence idea: the physical imprint, the body that carries out the computation will be part of the problem and part of the solution. Maybe that was the message of the talk, that I'm finally getting.

Engineering, not computation?

I heard Noah Cowan of JHU give a great talk today. He talked about having an engineer's perspective, a systems perspective, applied to scientific questions in biology. There is so much to be learned about the neural control of animal behavior, and he and colleagues are figuring out how animals close the sensorimotor feedback loop. It turns out, for example, both theoretically and experimentally that cockroaches can't be using simple proportial control when they are running at 1.5m/sec along a wall, but a PD model does in fact predict pretty well the roach motion. Or there are these weakly electric knifefish that really like hiding inside tubes. If you move the tube around, it will look like the fish is attached to it with a spring as it follows along. And it turns out that a harmonic oscillator (spring-mass) mechanics model in fact correctly predicts the behavior at various frequencies of moving the tube around, whereas a kinematic model does not.

As I payed close attention to the controller-plant-feedback diagrams, the second-order differential equations and the low-pass, high-pass and band-pass filter discussions, I noticed that all of these useful tools have nothing to do with computation or a computational worldview. The engineer has a formidable toolbox for use in the science of neural control; the progress that can be made is remarkable. But what can a computer scientist bring to the table? What models, what theoretical tools?

It's pretty obvious to me that computation won't be helpful in figuring out the science of animal motion, in all its graceful and fast glory. So what kinds of biology questions might we address armed with our understanding of data structures, algorithms and complexity? Ideally, questions beyond proving this or that purportedly intelligent activity is NP-hard? My hunch (and I'm betting my research on it) is that our computational tools will be best employed in asking and answering questions at a level above that of the neural control of one organism. Instead, we can study the interactions, communications, and group-level behaviors of many organisms.

Bayesian models of human cognition notwithstanding.

I thought I was joking

But actually, today and tomorrow, the Neukom Institute for Computational Science at Dartmouth College is hosting a conference on The Human Algorithm. I wish I knew about it in time to go there! Speakers include Daniel Dennett, Patricia Churchland, Marc Hauser and others for an impressive line-up who for sure will reveal the exact steps to be taken by any machine longing to be functionally equivalent to a human being. Free will and legal responsibility included.

I'm only half-joking. Maybe they will publish the proceedings.