Hacker News Reader

Comment by ravenstine | Hacker News Reader

You can calculate all the probabilities you want all day and night, but there's still no evidence. Until we have a frequency of how often life occurs, rather than a single occurrence, speculated probability doesn't mean anything. Especially when we can't even create life in a lab. (If I remember correctly, we've been able to create amino acids using early-Earth conditions, but we haven't really been able to create something like RNA.)

The Fermi Paradox hasn't been "dissolved" in any way. It's a response to the idea that the universe is teaming with life. The paradox remains unchanged until we find a credible signal, or we create life in a lab under naturalistic conditions.

View on HN

You might want to look up the meaning of paradox.

The Fermi Paradox has been dissolved in the sense that there is no longer any paradox; that is, there's no contradiction between what we've observed (no life so far) and what some people expected from doing a rough calculation.

So yes, we still don't know if there's other life in the universe or not and there's no new evidence. But, at least for some theorists who accept this answer, a logical puzzle is no longer a puzzle.

Yes, it's not a paradox... to someone who accepts the Drake Equation as illustrated in the article.

No, it is not a paradox to anyone anymore. If you accepted common terms before, you had a paradox because your calculation yielded millions of civilizations and we've observed none. If you didn't accept those terms in the first place, there was no paradox because either you had no opinion of the matter or your calculation yielded close to zero to begin with. Since that fits the observation there is no paradox there either.

well summarized, should tag tl,dr

This is addressed in the paper.

They use Bayes to update their estimates based on the lack of evidence so far (which is, itself, evidence for lack of advanced stellar civilizations).

It has been dissolved in that they've pointed out a fundamental error in the logic that leads to the paradox, and have used statistics and a review of available literature to illustrate this error.

How could you possibly use bayesian updates to calculate the true priors based on only one sample? That's like saying "I have calculated that the sun will never rise again" after seeing it set for the first time in your life.

Your analogy would more accurately be, "it's been dark all my life, so the sun's never going to rise again."

The paper goes into this starting on page 12. After establishing estimates based on a variety of factors, they then adjust the results by pointing out that "there were no Type III Kardashev civilizations using more than 85% of the starlight in 10^5 surveyed galaxies" and that "the lack of observable settlement provides a strong bound on extraterrestrial intelligence".

They proceed to offer approximations for Bayes conditional probabilities with different formulae for no stars having been observed to have civilizations, no civilizations within a detectable distance, and no settlements within nearby spacetime.

After putting all of this together, they conclude in section 4.0 that, "The Fermi observation thus raises our 52% credence for being alone in the galaxy to somewhere between 53% and 99.6%, depending on the type of evidence considered. It likewise raises our 38% credence for being alone in the observable universe to somewhere between 39% and 85%."

Essentially, they're saying that if the probability of detection is some value if there are intelligent civilizations in the observable universe, then how does it change the model that they haven't been observed? I am not a knowledgeable enough mathematician to judge whether this is an appropriate use of Bayes or not, but I would expect that someone would have already pointed it out if it weren't.

They then conclude section 4.1 with, "...unless one assigns a very strong prior to [a systematic failure to detect observable civilizations], our qualitative conclusions still hold."

Absence of evidence isn't evidence of absence - especially not when dealing with entities and civilisations about which we not only know nothing, but about whose properties we can barely speculate.

The only thing Drake can possibly estimate is the number of civilisations similar enough to our own to be recognisable.

There's an assumption that every civilisation will pass through a stage where can recognise it - but of course there's no good reason to be confident this is true.

So in fact the L term is unknowable.

There's also the fact that not all civs will transmit, but some would be visible to better spectroscopic tools. Generally, as our technology evolves you would expect our detection skill to increase too.

This isn't included in the original equation.

> Absence of evidence isn't evidence of absence

Yes it is. You may be confusing 'evidence' for 'proof', but here's a proof that absence of evidence is indeed evidence of absence. (How much? That depends.) http://kim.oyhus.no/absence/index.html

Honey, did you leave the lights on downstairs?

Well, I don't see any evidence of light, but that isn't evidence for the absence of light. So I guess I just can't know.

> There's an assumption that every civilisation will pass through a stage where can recognise it - but of course there's no good reason to be confident this is true.

That's the inherent problem with only having one example of something. For all we know, our current way of being is fundamentally closing off our perception that would allow us the knowledge that intelligence exists beyond our own. Even our own thought patterns could be fooling us.

For example, we could very well visit a planet in another star system and be like...

"Wow, this whole planet is covered in rust. What a wasteland! Let's get out of here."

And we might not even think that the rust is actually an intelligence. We might figure out that it's living, after years of stomping on it or trying to mine base elements out of it, but we might never know that it's actually much more intelligent than we are because it doesn't have the innate human drive to spread out, conquer, explore, etc., all of which are not necessarily indications of intelligence.

Well, no. Amino acids seem to happen no matter what they try in the lab. Nucleotides have happened many times, starting in the '60s, as have polypeptides. Complex organics are nearly inevitable. There have even been RNA oligomers formed on a mineral substrate under conditions that probably existed in the Hadean[1] - so basically lab-grown prebiotics. An awful lot has happened since the Miller-Urey experiment. That doesn't mean that life as we define it is inevitable, or that it will survive all of the various attempts the Universe makes at eliminating it, what with all of the huge rocks and snowballs being thrown around and oozing and exploding lava and such.

The most basic building blacks just happen. The minor assemblies seem to be easy. The smallest key pieces seem to involve a bit of luck in the beginning - it may take a handy bit of inorganic scaffolding, some weather and something like tides to build the first self-replicators, but one you have them, the scaffolding isn't needed anymore. Getting from there to proper cells - replicating, metabolizing and homeostatic - takes a considerably bigger collection of chance meetings of chemistry and environment, which may take just a bit longer than we've been trying so far, and a much bigger lab. Even if replicators are common, "real" life might not be.

And then there was a little more than a three-billion-year wait for cells to stay stuck together and cooperating in the only successful experiment we know of. Multicellular life may be rare, even if life is common. Tool-using, highly intelligent life that has harnessed fire to its will doesn't necessarily follow, and that's pretty much a sine qua non of a civilisation capable of interstellar communication, let alone interstellar travel.

No matter how easy the first 500 million years or so of the formation of life on Earth may be, the assumption that what happened over the next 4 billion years is somehow a straight shot is one hell of an assumption. Almost nothing happened for about three billion of those years, and the catalyst for multicellularity may have been an exceedingly rare, even unique, set of circumstances.

The universe, the galaxy, may indeed be teeming with life, but it's much, much more likely to be little green slime-covered mud mounds than little green men.

[1]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1664692/

This is fascinating. Thanks for your reply! My information was pretty outdated.

Until we find life elsewhere, all we can do is speculate. No matter what happens, unless we find electromagnetic signs of advanced lifeforms, we'll be stuck being only able to speculate for a long time.

I do think it's a fair inference that there can be life elsewhere, and even that there is life elsewhere.

Also, the Fermi paradox is about advanced lifeforms. The universe may be teeming with primitive life forms. If we can't find any primitive life elsewhere in the Solar system, then we might not find out until we develop interstellar travel, and even then, "find out" would mean on time scales longer than a single lifespan.

Pointing out that we can only speculate seems... given that this is obvious... unnecessary. Perhaps you meant that we should not speculate?

Even if we create life in a lab under naturalistic conditions, that doesn't answer the question. How often do those conditions occur on planets in the universe? We need an actual number. How often does life arise given those conditions? We need an actual number for that, too.

Also, what if there are other conditions life can start under?

And what are they? And how probable are those conditions? And how often does life start under those conditions?

And what is life? Does it have to be DNA-based replicators? Just carbon-based? Matter-based?

Reply on news.ycombinator.com