Strange matter.

This is from Peter Singer's review of Richard Posner's new book Catastrophe: Risk and Response":

High-energy particle accelerators, used by physicists to investigate the fundamental laws of nature, could produce particles that create hyperdense ''strange matter'' that in turn might attract nearby nuclei, thus growing larger and attracting ever more nuclei, until the entire planet is compressed into a sphere no more than 100 meters in diameter...

The official risk-assessment team for one of these accelerators, at the Brookhaven National Laboratory, offered a series of estimates, one of which puts the annual risk of a disaster at one in five million. That seems a very small risk. But since the disaster would kill six billion people, that estimate gives it an expected cost of 1,200 lives per year. Even if the risk is estimated more conservatively at one in a billion, it has an expected annual cost of six lives. Would we build such an accelerator if we knew that six people would die every year in which it operates?

I don't understand the purported helpfulness of converting the problem of destroying the entire world into an annual death rate calculated along with the low probability. One in five million already sounds like a ridiculous chance to take. Can somebody remind me why we let scientists fool with particle accelerators?

UPDATE: James Bellinger, an associate scientist in the High Energy Physics Unit here at UW-Madison, emails:

Cosmic rays hit our atmosphere with a spectrum of energies: the higher energy, the rarer they are. We've seen evidence of energies so high that no earthly accelerator planned so far comes close to matching the center-of-mass energy you get when one of those cosmic rays hits a nucleus. Given that the earth (and the sun, or anything else in our solar system for that matter) hasn't imploded yet, you have to conclude that the risk is pretty small that undiscovered strange matter physics is going to kill us.

And some people have looked for evidence of "strange nuclei" in cosmic ray interactions. The last time I looked at this was back in 87 or so, but a quick googling about this morning finds a Japanese team which claims to have found a couple of odd events (PRL 65 p2094, 1990) and an upper limit on rates from another group. If the Japanese result is correct, then we at least know that strange matter doesn't destroy balloons . . . And there's this.

So please, let us keep playing with accelerators :-)