Life emerged very early in the geological history of Earth. The probability of this happening is greater under the "P(life|conditions)≈1" model than under the "P(life|conditions)≈0" model. Our historical observations give the "P(life|conditions)≈1" model greater likelihood. It's not exactly proof, but it's enough to convince me.
So that brings up the question, if life happens so readily given the right conditions, how come it only happened once on earth? And I don't buy the explanation that one form out-competed all other forms -- given the diversity that we have now, there can be millions of forms of life all existing at the same place/time. But all instances we have now originated from one single biogenesis event.
So, how do we know that there was only one biogenesis, and not multiple ones that happen to resemble each other (i.e., that DNA-based life is the only thing that ends up working, and it occurred multiple times)? From what I gather, there are many markers in biological molecules, such as the direction certain molecules form (left vs. right, etc), and that multiple biogenesis events would have gotten these in random positions (i.e., your heart on the right side instead of the left side of the body).
> given the diversity that we have now, there can be millions of forms of life all existing at the same place/time.
Sure, each niche favors a different specialist, but why do you think that whichever process led to the emergence of life will be any good at producing competing specialists? I'd think that evolution of existing organisms would be infinitely better at this task, as existing organisms have both a numerical advantage and the strategic advantage of DNA (a "playbook" of previous successful strategies to draw inspiration from).
Very few niches are available for the emergence of life, even though there are clearly many available for the evolution of existing life. The available niches require a level of sophistication that spontaneous process cannot achieve but that evolution can achieve (e.g. photosynthesis).
Actually, given that our star is a third-generation star, you could even ask the question : why wasn't there life in our very own solar system before earth even existed ? Before sol started burning or even coalesce ?
The first generation stars wouldn't have had anything but hydrogen clouds surrounding them, but second generation stars would have had similar amounts of other elements to what we have today floating around them.
Given that "our" biogenesis event happened only the second time it could happen (here and in billions of other solar systems in our galaxy), where the hell are the second generation societies ? Didn't they survive the supernovas ? (possible, I suppose, but not exactly hopeful for our own chances of spreading across the stars). Was there some kind of large scale disaster ? But the question is worse than that, because they should have been spacefaring ... if they had anything like our numbers of satellites, we should have been able to find something, somewhere, right ?
Not strictly relevant. If your hypothesis is extreme, a single observation can make its posterior probability small. For instance, suppose that Alice gives you a coin and says "it lands on heads 99% of the time" (hypothesis A) but Bob calls BS and says "it looks like a regular quarter, so it should land on heads 50% of the time" (hypothesis B). You assign prior probabilities of 50% to A,B because on one hand you can't imagine how anyone could engineer a coin to land on heads 99% of the time, but on the other hand Alice is a great engineer, so you don't know what to believe. If you flip the coin a single time and get tails, your experiment strongly supports hypothesis B over A. OTOH, if it landed on heads, A would be weakly supported over B. We started with an even prior, so the posterior probability of A,B are proportional to their likelihoods: P(A|t)≈2% and P(B|t)≈98%, while P(A|h)≈66% and P(B|h)≈33%.
Let's say that a typical planet remains habitable for 8b years and it takes 4b years from the emergence of life to the emergence of intelligent life, so according to the anthropic principle life had a 4b year window in which to emerge. Under the model where P(emergence)/time is tiny, the probability of life emerging in the first billion years is ~.25. Under the model where P(emergence)/time≈1, the probability of life emerging in the first billion years is 1. With even priors, the posterior of P(emergence)/time≈1 is 80% and the posterior of P(emergence)/time≈0 is 20%.
It's not scientific proof of anything, but it's enough to make me consider alternative explanations for why life is rare.
