Tuesday, September 30, 2008

ITER vs The Stone Axe

Stephen Strauss takes a look at big science and comes away unimpressed. He talks about two exhibits he saw. One for the $15 billion ITER (pronounced EATER - heh) and another about neolithic technology - mat weaving, pottery making, chipping stone axes.
At the recent European Science Open Forum conference in Barcelona, for example, I was strolling through exhibits aimed at — please don't gag — science outreach. The underlying theme of all these displays seemed to me to be: since their schooling actually teaches many ordinary people to be discomforted by — if not to actually fear and loath — science, let's see if we can't do something in these venues to get people to hate science a little bit less.
And why do people hate science so much? Well it is hard to understand and requires a lot of complicated math and difficult concepts. I'm pretty good with that sort of thing. I understand Einstein but the math is beyond me. String Theory? Fuhgeddaboutit. So how about neolithic technology?
Right across from ITER was an exhibit in which a group of paleo-archeologists had set up a display to show the technology of the past in operation. So you had a guy sitting cross-legged, banging away at a rock to make a hand ax. Chip, chip, and chip. You had someone else weaving plants together to make a mat. Weave, weave, and weave. Someone else was taking clay and making a pot. There was no placard asking: Hand axe making, will it always be 40 years away? There were no critics of the effort calling it a huge waste of national resources.

So what does the juxtaposition of the two very different demonstrations of technology tell us about disbelief?

To begin with, the ITER project and all hugely expensive big science efforts — think the International Space Station, think Large Hadron Collider, which recently has received a tonne of press — aren't like making hand axes. I looked at the man diligently chipping away and realized that the price of his failure wasn't very high. So what if it turned out the rock type you made axes from wasn't strong enough to chop wood? You simply went back and made axes from something else until you got an ax that worked.

And you, in this case, would simply be some intrepid carver and not some large part of the Paleolithic science world.

On the other hand, if ITER fails, it is massively unlikely there is going to be another effort to correct its errors. Research on its level is simply too big and expensive and time-consuming. But what if it succeeds — but only kinda? What if its results show that you can produce energy, but that it is 10 time times more expensive than energy from other sources? What if figuring out how to make that equation more favorable will require at least three iterations of ITER?
So how should we be thinking about such projects? A little differently to be sure.
What you put in place with these vastly expensive research efforts is a "can't afford to fail" paradigm. Unlike trying to find the best plant material to weave into a mat, ITER, the Large Hadron Collider, etc., must succeed on first go-round. With ITER, there is no second kind of rock to be chipped away, no other plants to be woven, no different type of clay to be baked into a plate.

And that's what I so disbelieve about it. It's not really experimental science; it's risky, we-can't-fail, all-or-nothing science and I would respond to that paradigm with the wisdom of stone axe makers.

Sometimes your research should be based not on how glorious success might be, but on how little you will have lost if you screw up.
So what should we be doing about fusion? Lots of small "understand the science" and "proof of concept" projects. Say 100 two million dollar efforts. About 10 twenty million dollar efforts based on the successes of the two million dollar jobs. And one or two two hundred million dollar efforts based on the promise of the $20 million efforts. Total cost of around a billion dollars a year when everything is fully ramped up. Nothing that is too big to fail and nothing where testable results are fifteen to thirty years off.

Of course I have my favorites. Here is one that I described in the Fusion Report of 29 August 2008.