Hacker News Reader

Comment by titzer | Hacker News Reader

None of this probabilistic reasoning matters in the face of the anthropic principle. If there are infinite universes, we could easily be in the mostly stupidly improbable one possible, since we are not observing (and cannot observe) all possible universes, let alone an average one. This why a whole ton of speculation is completely meaningless, because we clearly live in a stupidly improbable universe that is already "fine tuned"--and that's not a problem. It doesn't mean there has to be God or anything. In fact, the settings of our universe imply nothing.

View on HN

Indeed we might never be able to measure the global normalization of the likelihood function of life, but that doesn't mean we can't investigate it's local properties. In fact, I think we can learn a lot from doing just that.

This article's probabilistic reasoning seems interesting, because it suggests it is very unlikely we will find other random intelligent life within our observable universe. As opposed to another sense of the anthropic principle, where our universe is fine tuned to an extraordinary degree but the chances of finding other intelligent life nearby is higher.

> None of this probabilistic reasoning matters in the face of the anthropic principle... In fact, the settings of our universe imply nothing.

What a sad anthropic principle that would be!

In fact, you can use the fact that humans exist to put (loose) constraints on this universe's parameters. For example, clearly the average temperature must be below one billion Kelvin. Making that kind of argument tighter and more precise actually leads to interesting results: bounds on spacetime black hole formation rate, limits on densitiy of dark energy, etc.

The authors in OP's post take an interesting variant of this line of reasoning: If the universe is "really bad" at producing long strings of RNS, then humans must just be a fluke. In that case, maybe most life in the multiverse exists alone in its own observational bubble. They caculate with their model that self-replicating DNS is indeed hard to replicate. This also means that finding aliens would provide strong evidence against their model.

The weakest form of anthropic argument is the tautological aspect you mention, but I wouldn't cut it so short.

> to put (loose) constraints on this universe's parameters.

I was referring to the fundamental constants of physics. A lot of argumentation has happened around whether they are "fine-tuned". In the sense that a small change has a radical effect on matter and spacetime, to the point where chemistry would not work at all to make any kind of biological entities, YES, YES, absolutely, our universe is fine-tuned for human life. Or vice versa. Human life is fine-tuned for exactly this universe, planet, climate (!), and time period!

There are several traps that philosophers and cosmologists keep falling into, primarily around probability. Clearly we are not in an average universe, from the very get go. We have no idea how these constants get set, what set them, if there was a universe before that had different constants, if universes elsewhere have different constants. Even if we did know that other universes exist, we have no way to reason about the probability of our universe--we don't know the distribution. If we were to start positing that the constants are randomly chosen, then we're in a very special universe indeed. But somehow, when it comes to these constants, the "probabilistic thinking" just gets forgotten. Philosophers just blink and look away. These constants are special as hell.

The next issue is that there is a huge philosophical problem in assuming that probability is even a thing, because every rando weird thing that has happened has fed into the stupidly improbable fact of our existence. The assumption that we are drawn from some kind of distribution is just a philosophical axiom that goes unstated. Maybe we aren't drawn from some probability distribution but are instead linked in causal chains impossibly long and intricate that using probability has a tool becomes meaningless. To understand what I mean there, just recall that the dinosaurs ruled the Earth for over a hundred million years. They were wiped out by an asteroid with a stupidly improbable chance of impact. The mass extinction that followed was so brutal that literally every single thing that survived pretty much made it by the skin of their teeth--including mammals. So every Tom, Dick, and Harry rabbit that managed to eek out survival on the blasted plains, carrying their painstakingly tuned genetic code, mattered. To the extent that many whole genera of life today depend on literally just a few individuals having survived this cataclysm. The Earth didn't mint new kinds genomes from nothing, they grew and branched off what survived. And it wasn't luck...they were better adapted to the new hellscape. Or was it? Sliced as finely as single individuals, we might be here because a single proto mammal survived a harsh winter or leaped out of the jaws of a predator at just the right time. Put odds on that one.

But speaking of "luck", it wasn't "luck" that the asteroid hit at that moment. It hit because of orbital mechanics, a complex but completely predictable mathematical certainty of orbits and periods. Pretty much all the macro events that we observe as being to have led here to this moment with these funny barely space-faring apes yammering at each other through HN are deterministic, not stochastic events. And the next thought that pops into your head on how to reply won't be stochastic either.

Reasoning about our universe using some Drake-equation-esque probability floundering is a complete waste of time. We got here by history and this whole freaking show is one giant example of selection bias. Average universe, my ass!

The probability for one asteroid hitting earth is low, but the probability that some asteroid will hit in a billion year time window is extremely high.

The specific asteroid that hit the Earth was not produced by a stochastic process, but a deterministic one. You are mistaking the tool (probability modeling) for reality. In reality, had that particular asteroid not hit at that time, yes, sure one would eventually have, but it might have been tens or hundreds of millions of years later....that a different asteroid hit, at which time the ecosystem of non-dominant species could have been completely different, and mammals would not have come out on top (and led to us). So when you actually look at reality and the casually-related events, you have to keep taking on massive amounts of zeroes to come up with probabilities for how things happened in exactly just the right way (this way)--we are talking thousands, maybe hundreds of thousands of zeroes, like 10^-1000000. That's exactly the problem with probability going wonky; so many things are not independent random variables. The whole analysis is wrong.

But that's just part of what I am saying. The odds are low, yes, but everything is distorted because of survivorship bias. Since the odds of our reality looking (exactly) like this are astronomically low, it is by definition not the case that we are in an average scenario, we're in a super rare scenario; but moreover, we are survivors in that scenario, and literally everything we look back at is hugely biased by the fact that we exist. So we are absolutely not in an average case, because the selection bias of surviving is inescapable.

Put another way, there could be a process that rolls a die every second and 99.99% of the time, destroys the universe. In the .01% of time, we survive and don't even notice. We keep perceiving those .01%, and multiplied together, the odds of surviving even a single minute are astronomically low. In the average universe in this scenario, everyone is dead! But we can't perceive universes in which we are dead; we wouldn't even be able to discover this process exists! So the whole business with universes randomly chosen from a distribution is just a bad model, as clearly our existence cannot be explained that way. Philosophers and cosmologists keep assuming that this is going on and that we must be in some average scenario. Uh, no.

Anthropic principle makes sense only for the past, not the future. We live in a universe with certain rules that make life possible, but there is no guarantee for the future. So why do the rules stay the same? The Bolzmann brain argument seems more probable.

what do you mean by guarantee for the future, why do you think that our universe have "guarantee for our future"?

Perhaps I should have phrased it differently to make it more clear:

If you make the assumption that our universe is drawn out of an infinite set of universes (saying we're here because of the anthropic principle), then the likelihood that life is possible in the future approaches zero, because the anthropic principle only says something about the past.

> anthropic principle only says something about the past.

and it says nothing about the future, so I don't know why you think "likelihood that life is possible in the future approaches zero". If anything the possibility in the future correlates with it in the past.

Also we are not concerned with life that will start in the future, when we talk about extraterrestrial life, we're talking about now or in the past.

> and it says nothing about the future, so I don't know why you think "likelihood that life is possible in the future approaches zero".

Because usually the assumption in the context of the anthropic principle is that the likelihood of life is small, but the anthropic principle makes it appear as if it's 100%.

There is no evidence that there are infinite universes and indeed the belief in such a concept absent any evidence beside "nothing says there can't be" by scientific people is a religious belief on the order of there being a god except with less elf honesty.

We have never detected an alternate universe and we don't even have math that makes it probable. It's just a concept. The same way "there's no reason entropy can't run backwards according to the physics" doesn't stop us not having any evidence for backwards running entropy or negative mass or a million other things that might conceivably exist but have no real evidence but sound cool.

The most philosophically and mathematically consistent interpretation of Quantum Mechanics is that there are "many worlds". This is a very strong hint (but not proof, per-se) that there isn't a singular, unique Universe in the classical sense.

In other words, there is a continuum of possibilities where all scenarios consistent with the laws of physics are played out.

Note that this is not evidence (not even circumstantial) that there are other universes with different laws of physics. Just universes with different states or boundary conditions.

Still though, if MWI is true, then there are effectively infinite possibilities in which the chance of life occurring somewhere approaches certainty.

> The most philosophically and mathematically consistent interpretation of Quantum Mechanics is that there are "many worlds".

Why is "many worlds" the most mathematically consistent? (I won't bother asking for the philosophical bit; I'm sure it's too long for an HN post.)

We have no evidence that we are randomly distributed over some population of universes, but that doesn't stop misguided philosophers from assuming that. I didn't assert infinite universes; rather the opposite. It's reasoning like this that is asserting other universes. One thing is certain; ours is super weird and has definitely produced self-aware apes in a long series of evolutionary accidents on a single planet.

What's more interesting is whether our universe can be described by finite information (closed formulae or not, even a stupidly huge but finite trace of all quantum events would suffice). If it's finite information then its description is encoded in every transcendental number in mathematics--which is an infinite number, including the number pi.

Don't trifle with infinity. I wasn't.

Having any belief at all, none, one, many, infinite, all falls into the same category. We don't know.

To me it seems like the additional universes in this context serve a purpose similar to (my understanding of) additional dimensions in math/physics. It allows you to navigate around singularities.

> Having any belief at all, none, one, many, infinite, all falls into the same category. We don't know.

If you have no evidence for it, your belief in a multiverse is religious not agnostic.

>To me it seems like the additional universes in this context serve a purpose similar to (my understanding of) additional dimensions in math/physics. It allows you to navigate around singularities.

This is additional dimensions in physics, and we just don't have any actual evidence that a multiverse exists. We have a some suppositions that say 'well there's nothing to say it couldn't exist' but no evidence that it could.

You could view that as 'useful' but you can't judge its truthfulness or whether reality reflects such a 'purpose'. Either something exists or it doesn't, and a multiverse is a strictly religious belief at this point.

Reply on news.ycombinator.com