Why I Think We Are Alone in the Universe
(And No, I Don’t Believe in Creationism)
I think we are alone in the Universe.
Many people will argue that, given the sheer size of our galaxy — let alone the observable universe — and the estimated high number of life supporting planets, it’s a near impossibility that there is no other life out there.
That’s an exciting prospect and I was arguing the same way in the past, but I’ve changed my mind.
The reason I changed my mind is the following: As mind-blowingly enormous the observable universe and even our galaxy alone might be, probabilities for the occurrence of possible events can be just as mind-blowingly miniscule (in fact, while the observable universe is finite and has an upper bound, the probability for the formation of life happening might be arbitrarily small) and thus the size of the universe or the number of life supporting planets really is irrelevant so long as one doesn’t also provide sound reasoning on the probability of life actually forming on those planets.
As it comes to this reasoning, I came across two types:
- Likelihood estimates based on actual explanations for how the formation of life can take place (bottom-up)
- Likelihood estimates based on the fact when/where/that our particular formation of life took place (top-down)
The first type seems to categorically yield extremely small probabilities, while the second one yields large probabilities (often 1 or within one or two orders of magnitude of it).
Unfortunately I found all of the arguments of the second category to contain logical errors, which I’m going to show.
The Drake equation is often used to compute an estimate for the number of other civilisations in our galaxy whom we should be able to communicate with:
The different factors are:
R*: mean rate of star formation
fₚ: fraction of stars that have planets
nₑ: mean number of planets that could support life per star with planets
fₗ: fraction of life-supporting planets that develop life
fᵢ: fraction of planets with life where life develops intelligence
fᶜ: fraction of intelligent civilizations that develop communication
L: mean length of time that civilizations can communicate
Depending on what numbers are used for the different factors (only the first three of which we can base on empirical evidence), a huge range of possible numbers results. The original estimates result in a range of 20 to 50'000'000 other civilisations in the Milky Way galaxy alone.
Based on these big numbers and in turn the apparent high likelihood of extraterrestrial life in the galaxy, the Fermi paradox arises, which I do not want to go into however, because I think the premise that life should form easily when the conditions are right is wrong in the first place (so I won’t base my argument in any way on the Fermi paradox and as such will also not go into the other factors of the Drake equation).
The reason I don’t agree anymore with most estimates based on the Drake equation (except for maybe the Rare Earth Hypothesis), is that the optimistic ones — which are the ones responsible for the currently popular opinion among scientists of there being a lot of extraterrestrial life out there — often assign a value of 1 to fₗ, the probability that a planet capable of supporting life also eventually will form life, baselessly assuming that when the conditions are right and the required elements are available, life will form with absolute certainty (if not 1, they’re making some other high a priori assumption of this factor).
To me this seems untenably optimistic if no solid explanation is provided of how self-replicating molecules — or any other chemical process that is capable of meaningful information processing and evolution — may form out of a sea of mere building blocks.
What we do know for a fact is, that life IS possible (otherwise you wouldn’t be reading this article). But why should we assume it to be a certain consequence of the conditions simply being right?
Usually the mediocrity principle is used to justify that this was a sane assumption. Earth is a typical rocky planet in a typical star system, located in a non-exceptional region of a common barred-spiral galaxy. Thus, so the argument goes, as we find ourselves in a non-exceptional place in the universe, by application of the mediocrity principle, it’s most likely that most of the universe looks very similar and hence it should be teeming with complex live.
But here lies the first clear error in reasoning: The mediocrity principle only applies to a draw at random, but in our case there is an obvious pre-selection applied (only draws that actually form life are counted) and thus anthropic bias applies to the observation that we exist: The question can only be pondered over in a universe in — and on a planet on — which life actually has formed. Thus, the mediocrity principle is clearly not applicable (I will apply it later in a more appropriate context though to show the opposite, that in fact it’s very unlikely that extraterrestrial life exists in our observable universe).
Another argument that I came across multiple times, which is in favor of a high likelihood for life forming on suitable planets, states that life on Earth formed very early in its history (already a few hundred million years after formation) and concludes that life therefore must form pretty easily.
This however is another logical error based on selection bias that can easily be uncovered using an analogy:
Imagine you‘re flying with an airplane for the very first time and you have absolutely no knowledge about the safety of airplanes. About 30 seconds after a bumpy take-off the plane goes into a nose-dive and crashes into the ground.
As the airplane explodes around you, you‘re making three observations:
- Your plane crashed.
- As far as you can tell, this is a very regular plane, it isn‘t in any way special compared to other planes.
- Your plane crashed very early into the flight.
From these three observations as well as your understanding that a plane must be flying for a crash to be possible at all, you conclude that airplanes must crash very easily once airborne and thus there must be crashing airplanes everywhere.
This is obviously an erroneous conclusion, because these observations don‘t actually tell you anything about the underlying likelihood of getting into a plane crash (other than it not being 0).
Conversely however it isn‘t very surprising that you found the crash occurring very early into the flight, because (despite you not being aware of this) plane crashes — as rare as they are — are most likely to happen during take-off or landing.
So, you are subject to selection bias and your three observations give you little to no useful information about how likely it is that airplanes crash, they are however in line with the fact that the conditions for a plane crash are best during take-off and as such it’s not surprising that the crash you found yourself in happened soon after take-off.
Additionally it’s important to stress though that even if plane crashes were most likely to happen in the midsts of the flight and you were aware of this, it would still be a logical error to conclude from the crash happening during take-off that planes would crash with high likelihood.
The only way to come up with a reasonable estimate for the likelihood of a type of event for which we have only 1 observation, is by coming up with an explanation for how this event came to be and then reason over the likelihood of the event via its explanation.
When back-tracking evolution, it’s clear that there must have been a first replicator: The very first molecule capable of replicating itself. There are good theories about what that first replicator might have been: Most likely self-replicating RNA, maybe some precursor of it (PNA, TNA, ANA…).
While there is widespread consensus that there was such a first replicator (or a dual pair of replicating molecules), there doesn’t really seem to be any widely accepted theory of how abiogenesis, the formation of that first replicator, could have taken place.
Empirically our sample size for the occurrence of abiogenesis in the universe is exactly 1 and we don’t even have any evidence for it having occurred more than once on our own planet. So, concluding that fₗ=1 (or any other high number), without having a theory free of logical errors of why this likelihood should be so high, is exremely bold.
Without any evidence that there exists a viable chemical reaction chain which yields a self-replicator, rather than being very high, I think it’s more reasonable to assume the likelihood of abiogenesis happening by pure chance is rather extremely small, as shown by quantitative approaches to the problem. Let’s call that extremely small chance of abiogenesis happening once in the observable universe Pₗ.
Creationists certainly like to see divine intervention in such an unlikely chance materializing, in the same way as they see the fine-tuned universe as evidence for intelligent design: The probability that the constants defining our universe are exactly such that galaxies and stars can form is incredibly small. Looking at this unlikely ‘coincidence’ from the perspective of quantum physics however, in particular Everett’s many-worlds interpretation (an interpretation endorsed by a big portion of quantum field researchers), the occurrence of this very improbable event is rendered unsurprising however.
To me, the most plausible explanation I’ve come accross so far for abiogenesis happening in the primordial soup is found in Quantum Biology. The idea (which is in line with the standard model of particle physics) is that the base components that formed the first self-replicating molecule were in a state of quantum superposition.
While in superposition, the various base components are inhibiting all possible combinations simultaneously, a few of which describe valid self-replicator molecules.
Looking at this quantum superposition using the copenhagen interpretation, the “observer” that caused the replicator-yielding quantum wave collapse would be life itself (our existence). In my opinion it’s much more meaningful though to take a step back and look at this superposition through the lens of Everett’s many-worlds interpretation:
Among the infinitely many possible arrangements of this superposition (of which the many-worlds interpretation states that all happen equivalently, each yielding a new branch of the multiverse), most of them yield random arrangements, but a few of them yield a functional replicator. From that perspective abiogenesis indeed is in fact inevitable — not on each habitable planet of each universe however, but across the entirety of the multiverse.
We live in one of the universes (corresponding to paths through the tree-like structure of the multiverse) that contain the highly improbable event of such a self-replicator forming out of its abiotic base components. Since there is selection bias, we may wonder about that ‘coincidence’ — but from the perspective of the multiverse, it just means there inevitably exist branches in which life forms emerge and wonder how it can be that such an unlikely event occurred.
Doesn’t that mean there are also branches of the multiverse in which abiogenesis happens more than once? Absolutely, infinitely many actually.
HOWEVER — and now I’m going to apply the previously incorrectly applied mediocrity principle, without being fooled by anthropic bias:
The conditional probability that a second such event occurs within that branch of the multiverse in which we find ourselves (the one in which I’m writing this article and you are reading it) is Pₗ — extremely small — and for n additional such events happening, the conditional probability is Pₗⁿ. Thus by application of the mediocrity principle — which this time describes a draw from the set of all universes that contain one or more abiogenesis events — we are most likely to find ourselves in a branch of the multiverse that contains only a single abiogenesis event (as those are more numerous than those with multiple such events).
Note that you don’t need to subscribe to the many-worlds interpretation of quantum physics, as you arrive at the same conclusion if you interpret that event using copenhagen + weak anthropic principle.
TL;DR
My reasons for changing my opinion on extraterrestrial life existing in our galaxy (and even our observable universe):
- While the observable universe is mind-boggling enormous and contains an equally enormous number of planets that could sustain life, this immensity is completely irrelevant without an argument regarding a lower bound for the probability of abiogenesis happening.
- The observable universe is finite, while the probability of an abiogenesis event may be arbitrarily small.
- All top-down arguments regarding the probability of abiogenesis based on the fact THAT, WHEN or WHERE it happened in our history contain logical errors (usually some form of violation of the weak anthropic principle, which is a logical tautology)
- All bottom-up arguments regarding the probability of abiogenesis happening based on actual mechanics of which we do know could generate a first replicator yield extremely small probabilities.
- Since from the set of available arguments regarding this probability those in favor of a high probability contain logical errors and/or make more a priori assumptions, by application of occam’s razor the ones that make fewer assumptions and don’t contain logical errors are to be preferred (and unfortunately those are the ones yielding a very low probability).
What are the consequences of this realisation? Probably none that affect us directly, other than the resources we spend on trying to contact E.T. mostly being spent in vain.
It does seem to charge us with a higher responsibility though:
Should we ultimately destroy ourselves, we can’t take solace in the thought that life will nevertheless prevail in our universe and as such we should invest even more into becoming a multiplanetary species and into seeding other planets with life.