Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Thursday May 21 2020, @01:38PM   Printer-friendly
from the we-just-don't-have-anything-interesting-to-say dept.

New study estimates the odds of life and intelligence emerging beyond our planet:

We know from the geological record that life started relatively quickly, as soon our planet's environment was stable enough to support it. We also know that the first multicellular organism, which eventually produced today's technological civilization, took far longer to evolve, approximately 4 billion years.

But despite knowing when life first appeared on Earth, scientists still do not understand how life occurred, which has important implications for the likelihood of finding life elsewhere in the universe.

In a new paper published in the Proceeding of the National Academy of Sciences today, David Kipping, an assistant professor in Columbia's Department of Astronomy, shows how an analysis using a statistical technique called Bayesian inference could shed light on how complex extraterrestrial life might evolve in alien worlds.

"The rapid emergence of life and the late evolution of humanity, in the context of the timeline of evolution, are certainly suggestive," Kipping said. "But in this study it's possible to actually quantify what the facts tell us."

To conduct his analysis, Kipping used the chronology of the earliest evidence for life and the evolution of humanity. He asked how often we would expect life and intelligence to re-emerge if Earth's history were to repeat, re-running the clock over and over again.

He framed the problem in terms of four possible answers: Life is common and often develops intelligence, life is rare but often develops intelligence, life is common and rarely develops intelligence and, finally, life is rare and rarely develops intelligence.

This method of Bayesian statistical inference—used to update the probability for a hypothesis as evidence or information becomes available—states prior beliefs about the system being modeled, which are then combined with data to cast probabilities of outcomes.

"The technique is akin to betting odds," Kipping said. "It encourages the repeated testing of new evidence against your position, in essence a positive feedback loop of refining your estimates of likelihood of an event."

From these four hypotheses, Kipping used Bayesian mathematical formulas to weigh the models against one another. "In Bayesian inference, prior probability distributions always need to be selected," Kipping said. "But a key result here is that when one compares the rare-life versus common-life scenarios, the common-life scenario is always at least nine times more likely than the rare one."


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Informative) by theluggage on Thursday May 21 2020, @05:47PM (2 children)

    by theluggage (1797) on Thursday May 21 2020, @05:47PM (#997463)

    It's more like this...

    You have a bag containing a mixture of green and blue balls. You don't know if it contains

    (a) 10 green balls and 9990 blue balls

    (b) 5000 green balls and 5000 blue balls

    (c) 9990 green balls and 10 blue balls.

    ...And because this is math, sensible actions like opening the bag ad counting aren't an option. Bloody math.

    Regular probability tells you that the chances of blindly pulling out a green ball are, respectively, either (a) 1/1000, (b) 1/2 or (c) 999/1000 and - assuming you replace the balls - that probability will remain constant however many balls you pull, because each selection is an independent event.

    ...which is, of course, pedantically correct, but realistically nonsense because each ball that you pull gives you a clue as to the distribution.

    if the first ball you pull out is green then - although it doesn't prove anything - hypothesis (a) just lost a lot of plausibility. You'll still need a lot of draws to work out a reliable estimate for the proportion of green and blue balls in the bag (which is what the mathematically correct but stupid high school probability problem would be) but even one or two draws would make scenarios like (a) and (b) very unlikely.

    (Complete the subject to make a well-know phrase or saying...)

    As for extraterrestrial life, we've only pulled one ball so far and it's green (well, actually a rather nice blue, white and what tends to look like brown marble pattern but it's metaphorically green) but, honestly, that is enough to put the burden of proof (that we're on the outer rim of some cosmic bell curve) on those who deny the possibility of extraterrestrial life.

    The point of the Drake equation is that it puts long odds on each step or a a chain of requirement for planets with intelligent life, and demonstrates that - when multiplied by the rather impressive number of stars in the galaxy - you still get a significant number. It's only an (astronomically large) ball-park.

    It's also worth remembering that, back when the Drake equation was formulated, all we knew was that a was couple of stars might have been wobbling because they had planets - now there is a huge catalogue of exoplanets, and not just ones that were only detected because they were three times the size of Jupiter. So some of the wild guesses in Drake can already be refined.

    Starting Score:    1  point
    Moderation   +3  
       Informative=3, Total=3
    Extra 'Informative' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   5  
  • (Score: 3, Insightful) by hendrikboom on Thursday May 21 2020, @07:52PM (1 child)

    by hendrikboom (1125) Subscriber Badge on Thursday May 21 2020, @07:52PM (#997541) Homepage Journal

    What we have to deal with is the conditional probability *given* that the first ball was green, because if it hadn't been green we wouldn't be around to ask the question.

    -- hendrik

    • (Score: 2) by theluggage on Friday May 22 2020, @01:39PM

      by theluggage (1797) on Friday May 22 2020, @01:39PM (#997830)

      What we have to deal with is the conditional probability *given* that the first ball was green, because if it hadn't been green we wouldn't be around to ask the question.

      OK, I shouldn't have said "we've drawn a green ball" - because, as you say, anybody capable of asking the question must - by definition - be on a green ball so it's not really the same as happening to draw a green ball from the bag.

      The real point is that - until we have a few more data-points we're in the educated guess business - it's to soon to even go Bayesian on it and expect any real results. All probability is good for here is "thought experiments" that help decide which is the best "null" hypothesis. The possibility that we're the only "intelligent" life in the universe can only be disproven if ET says "hello" - the best we have is plausibility arguments as to why it is not the best hypothesis.