RCS > The most plausible model for the origin of life is that it requires
> 10^100+ chemical trials for the right combination of circumstances
> to occur.
Anders> So which is this model? Your claim seems not to fit in well with the
current thinking about autocatalytic networks of RNA-like molecules, which would have a much faster evolution. In fact, it sounds like a common creationist claim, based on the idea that the molecules must randomly bump into each other to form a cell or something similar, instead of the bootstrapping process of increasing complexity most paleobiologists subscribe to.
No, I'm ascribing 10^100 to the RNA autocatalysis scenarios. I find the RNA scheme unconvincing: There are too many "could-haves" and "maybes" and not enough experimental support. Dyson raises the objection (to an RNA-first world) that the pieces of the monomers (sugar + amino acid) are not both stable in the same chemical environment, so it's unlikely that they both accumulated enough to react significantly.
Dyson argues that 7 or 8 ingredients in the cycle should be enough, but I don't know of anyone attempting to put some chemical flesh on the bones of his theory.
The idea of a whole cell coming together at random isn't just 10^100 improbable; that's in the 10^10^10 range of unlikelihood. It's more likely that a Conway Life replicator would arise in random soup, a mere 2^10^6 unlikely.
We do have excellent evidence of the adaptability of life to extreme environments, and some earth microbes have undoubtedly been boosted into escaping the solar system. We shouldn't be too shocked if we find Mars or Ganymede has organisms sharing DNA sequences with E. Coli.
The creationists have a few good points, just as the opponents of infinitesimals forced more rigor on calculus.
Rich Schroeppel rcs@cs.arizona.edu