[bayesianhorse]

Setting p=1 is sort of frowned upon in the bayesian methodology. Bayesian inference has nothing to do with this discussion, because the results can only be explained by speculation, and have not been externally verified etc.

Constructing priors from the knowledge of what has not been seen before seems to me like an abuse of the methodology. You might as well say "It's impossible because it's impossible."

I do tend towards a terrestrial origin, but it's certainly a good thing to investigate further how these organisms got there, what are they doing and what sequences do they have...

[danmaz74]

Actually, to me, this seems like a perfect situation to apply Bayesian inference: we need to update the probability estimate of a hypothesis (there is life in outer space) based on new evidence (they found those particles through their experiment).

Anyway, I completely agree with your last sentence: It's certainly a good thing to investigate, I'm not saying they're charlatans. It's the "very likely" that I didn't agree with :)