I found this pop-science news article where scientist are seriously trying to use Bayesian reasoning to change beliefs of the field.



7 comments, sorted by Click to highlight new comments since: Today at 9:37 PM
New Comment

Here's the actual article on the arXiv.

I find it a useful working assumption that everything on physorg is either wrong or copied from somewhere else where it's presented more accurately or informatively. I don't think I've found any counterexamples yet.

Anyway, the article is merely making the observation that what's known about when life on earth emerged doesn't constitute enough evidence to make our posterior probabilities for abiogenesis very different from our priors. This isn't exactly world-shaking.

What are the better news sources that you can suggest?

I don't offhand know of any good general sources for science news. Perhaps there aren't any, in which case a useful algorithm might be to look on physorg, and then whenever you see something interesting go to whatever source they cribbed it from.

(I'm not sure whether your question is intended to suggest that if I can't suggest a specific better alternative then I shouldn't be criticizing physorg. If it is, then I disagree with the general principle. Saying "X is bad in such-and-such a way" doesn't incur an obligation to suggest something else that does what X is trying to do but doesn't have that flaw.)

Thanks i guess i just need to hone my check the source skills. The () part is not implied :)

The premise that we do not have the information to conclude that abiogenesis on planets similar to our own is common is correct, but I'm not sure that this is particularly changing much about scientists' beliefs in the field. If my experience is representative, this seems more like an articulation of what most scientists in the field already believe.

The linked synopsis seems to significantly overstate the case the paper is making though. Compare

So, they contend, deriving numbers from an equation such as that put forth by Drake, only serves to bump up our belief in the existence of other alien life forms, not the actual chances of it being so.

When taken at face value, some might conclude that such arguments hold no more logic than arguments for the existence of God, i.e. it’s more about faith, than science.


A Bayesian approach to estimating the probability of abiogenesis clari es the relative infuence of data and of our prior beliefs. Although a "best guess" of the probability of abiogenesis suggests that life should be common in the galaxy if early- Earth-like conditions are, still, the data are consistent (under plausible priors) with life being extremely rare, as shown in Figure 3.

Also, note that the linked synopsis claims that the Drake Equation drives an overly optimistic value of the likelihood of extraterrestrial life, citing the aforementioned paper, whereas the paper itself only attempts to establish the possibility of fℓ (the probability of life arising on a planet capable of supporting it) being low, making no criticism of the equation itself.

Thanks for clearing that up

[-][anonymous]10y 3

some might conclude that such arguments hold no more logic than arguments for the existence of God, i.e. it’s more about faith, than science.


That sentence was almost jarring--the author apparently has no idea what a Bayesian approach is all about.