Hacker News Reader

Comment by gpsx | Hacker News Reader

This article does a good job of covering a lot of different ideas on this paradox. I have a problem with the idea of "big numbers" in it though, like where it implies 500 billion billion is nearly infinite, or that one-in-a-billion is a freak occurrence. There is a saying that if enough monkeys pound on a keyboard, one will type Hamlet. This would take a lot of monkeys. If we take the odds of hitting one character to be 1 in 40 (which doesn't account for properly hitting the shift key to make a capital), factoring in the 130,000 letters in Hamlet (according to Wikipedia), we get the odds of a monkey typing it at 6x10^208267. This number comes from 40 keys and 130,000 letters. How about when we start taking molecular numbers of items (6x10^23) and start raising that to a large power? I don't think it is a given that the number of planets out there infer there should be other life.

View on HN

This is essentially the problem that I have with the Drake equation. It encourages--whether that's the intent or not--taking a reasonably known and very, very large number (number of stars) and multiplying it by probabilities/percentages, some of which (we think) we know with at least order of magnitude accuracy. But others we have absolutely no idea. And saying some of those events (such as the evolution of intelligence as we understand the term) has a 1 in a billion probability of occurring SOUNDS like a worst case for-the-purposes-of-argument plug-in. But it's really not.

Another example I hate is the probability of reality being a simulation thought equation. If these equations had some confidence interval based on how little we know it would be clear how meaningless they are.

We can't calculate the probability of life arising though because even if we did know how it arose here—which we don't exactly—we don't know all the possible different ways it can arise, which may or may not also be a number raised to a large power.

That is a good point. I agree with you. And people do have suggestions as to how life can be more probable than naively expected. But this answers the question why there is life at all as opposed to why we don't see more of it.

My bias is that life is highly improbable and that we will not find any other life in our universe. But I also think it is no miracle that we exist, as I will describe. Given that I am trying to say this in one paragraph, it may not sound well justified, but it is how I view the problem. I'll start with Schrodinger's cat. A cat is in a box with some radioactive source that can decay and kill the cat. You can ask if the cat is alive or dead - and the answer is both. I will ignore the next part (Schrodinger's whole point) about what happens when the human observes it. There are different interpretations about why the human thinks the cat is definitely in one of the two states as opposed to both. This has to do with the human's observation and is not important for now. The point is that the cat is in both states. The universe can be in multiple states, some with life on earth and some without. Every possible state that can exist, does. So no matter how improbable we are, we will exist in some version of the universe. And there would be lots of other life too, just maybe not in our version of the universe. So the probability of life existing doesn't impact whether or not we should exist, but it does impact whether or not we see other life. I don't think we will. But as correctly pointed out, we don't know the answer to this right now.

The universe may / be as great as they say / but it wouldn't be missed / if it didn't exist.

- Piet Hein

> My bias is that life is highly improbable and that we will not find any other life in our universe.

It's probable enough for you to be writing this. The only options are 0, one and many. We can rule out '0'. That leaves us to decide whether the chances of us being the only one are larger than the chances that we are one of many. Obviously the second one has more chance of being true the one where there is only one. But we like to consider ourselves to be special so most people will believe the 'one'. Just like the sun used to revolve around the earth, now we're essentially seeing the universe as revolving around us. The alternative, that we're not special is not compatible with a lot of our collective culture.

The point is that there are multiple universes, according to the quantum mechanics interpretation I subscribe to. So all of your options are true: 0, 1 and many. My position, which is admittedly no better than a guess, is that most universes by a far majority have 0. By virtue of me writing this, among other things, we live in a universe with at least 1. Again, my position would be that it is unlikely we have more than 1 in our universe.

What you are saying it true under the assumption there is a single universe.

I think you're trying to articulate is the Anthropic principle [1].

It basically states that the probability that "some" life exists is 1.

I'm not sure what the name of this next one is, but I believe it's a widely defended scientific principle that we're "typical" in some sense. This principle gets hard to justify cosmologically (it reduces to Occam's Razor locally), but it makes sense to me. It rejects very clearly some quirks like the Bolztmann Brain [2] argument.

[1] http://en.wikipedia.org/wiki/Anthropic_principle

[2] http://en.wikipedia.org/wiki/Boltzmann_brain

> Obviously the second one has more chance of being true the one where there is only one.

That is not at all obvious to me. You are essentially taking the odds of the odds of something. A statistician can correct me, but I don't think it works the same way as calculating the odds of something.

Like, it's not as if we have a roulette wheel where "1/10^24" is an option, "2/10^24" is an option, "3/10^24" is an option, etc. In that case, "many" outweighs "1," certainly.

But if the odds of abiogenesis are so unlikely that we happen to be the only life in the universe, it doesn't seem like those odds must necessarily be wrong simply because it's logically possible for them to be more or less likely than they are.

But even then, it's kind of a wash. For as much as it could be logically possible for the odds to be more likely than they are (which would allow abiogenesis to occur more frequently), it would also be logically possible for the odds to be less likely than they are.

The WAP accounts for high improbability by positing a multiverse, but (at least as I understand it) it doesn't predict high improbability, even if there is a multiverse.

Also, in a multiverse scenario, the improbable thing might not be the formation of life, but the fine-tuning of the laws of physics which allows the formation of life. In which case life might be relatively common.

If we assume that this tuning is not an all or nothing event then the larger the possible universe size the more likely it is that we lie within an region where the features are tuned just enough to let one intelligent life form arise.

You can think of the tuned conditions that allow life to be common to be a bullseye that is surrounded by tunings that are close enough to allow life to occur infrequently. If you were to pick a region at random out of the regions that contain intelligent life then it is likely that you would pick one of the surrounding regions not the bullseye. This is also an explanation for Fermi's paradox.

You must factor in that such a region would contain a low percentage of the total number of civilizations intelligent enough to consider the question. A well-tuned universe would be teeming with life - it is therefore more likely that we find ourselves in such a universe (which was the point of the GP). You can think of it, not as a bullseye, but as a gaussian spot.

It strikes me that the issue at play here is how one chooses to "sample" the anthropic principle, for want of a better term. When you say "pick a region at random", you assume a uniform probability distribution over the set of all universes supporting life (which incidentally is not possible if the number is infinite). But you don't say how to resolve the probabilities with universes containing multiple intelligent life forms - you do select randomly among those too? If so, shouldn't you rather be picking "randomly" from the set of all intelligent life forms in all universes in the first place?

Yes it all concerns the shape of the distribution. I have not assumed any any sort of distribution by the way as I have no means of estimating it.

It may be possible in the future to estimate how well tuned a universe could be made for the appearance of intelligent life and also how large the surrounding non-tuned space could be. In the mean time we can only speculate :)

This is one explanation, but for me fails two tests. The first (pointed out below) is that it is heliocentric - it's about us.

The second is that it avoids freedom of will and consciousness in the sense of self perception. It assumes that I am a zombie, and I deny that. I concede that others may be zombies, although I think that they aren't. My actions either navigate me from state to state or create new states. Physics does not account for this; physics is incomplete.

I'll leave occams razor on the shaving table, as it's a heuristic and I don't like it for logic - but infinite infinities fails occams too.

What does "freedom of will" mean to you? What you describe (actions move you from state to state) sounds like determinism to me.

To decide which state to move to. That I (and others) are agents that change the future of the universe by choice within the frameworks of possibility.

We can't, but we can assume that any process that arises from a series of unlikely chained events (where the likelihood of each event are not correlated) will most likely occur near to the latest time possible. For example, if you run a trillion parallel 10-step chained experiments over a year starting at Jan 1 where each step is very unlikely to occur so that only 100 reach step 10, then the last steps are all likely to occur in December. The less likely each step is (or the more steps involved) then the more the experiments that reach the final step will cluster towards Dec 31. If you have a group of such events then you can actually calculate the likelihood of the chain series.

Intelligent life in a large universe is a similar to this imaginary experiment. While we only have one observed data point, we do know how close the appearance of intelligent life on earth is to the last possible time that it could occur. Outside of our geologically recent anthropogenic boost to the CO2 in the atmosphere, the long term trend of CO2 concentrations in the atmosphere is down (this is due to the sun becoming warmer over time requiring CO2 to be removal from the atmosphere to keep earth in the habitable temperature range). We are in geological terms close to the point where any further removal of CO2 will mean photosynthesis will no longer work (plants are already CO2 limited) - maybe a few 10 millions years. In geological time this is very close to our Dec 31.

What does this all mean? Basically it appears humans have evolved close to the latest time point possible. While we can't calculate the exact probability of intelligent life arising in the universe from our one example, we can say that the evolution of humans is consistent with the hypothesis that the evolution of any intelligent life in the universe is very unlikely.

Why do you think that current CO2 concentration levels are important for intelligent life?

Slight adjustment in how body works can easily adapt organism to new CO2 levels.

I did explain it in my post, but maybe it was too short. Earth has been kept in the habitable temperature zone over the last 3.5 billion years through falling CO2 levels. Around 3.5 billion years ago the sun provided ~25% less energy than now and the CO2 levels were much higher. This is the green house effect. As the sun got brighter over time the level of CO2 in the atmosphere has fallen. For more on this see the slow carbon cycle [1].

The problem is that the level of CO2 in the atmosphere is getting very low (~0.2%). At this level plants are starved for CO2 - if it gets down to 0.01% then plants can't grow at all and the ecosystem needed to support the evolution of intelligent life would collapse.

1. http://earthobservatory.nasa.gov/Features/CarbonCycle/page2....

So some CO2 is needed for plants.

It does not make certain CO2 levels a requirement for intelligent life.

Change in temperature does not necessarily kill intelligent life either.

At least for the foreseeable future, we still need plants and natural ecosystems to survive on Earth.

Photosynthesis is required for the evolution of intelligent life. The point I was making is that intelligent life has evolved on Earth pretty close to the latest time possible.

C4 photosynthesis is an adaptation to higher temperatures and lower atmospheric CO2. And C4 is a very recent innovation (25 to 32 million years ago, says Wikipedia).

But yeah, I don't know if even the C4 plants could handle 0.01% ambient CO2.

Yes C4 photosynthesis is an adaption to falling CO2 levels in the atmosphere. It does come at a cost which is why in areas that are not water limited C4 plants have not out competed C3 plants.

> CO2 in the atmosphere is getting very low (~0.2%).

We are currently at about 400 ppm (parts-per-million), which would be 0.04%. http://co2now.org/

About 20 000 years ago, the lowest was maybe ca. 200 ppm.

I've always thought, given the right circumstances, life will always be formed. Instead we should be calculating the odds of life NOT forming e.g. the half-life of staying-dead.

Our planet has carbon, oxygen, water, the right temperature, a moon that sloshes it all around, rotation. Every grain of sand on every beach is going through a different cycle of heating/cooling + wetting/drying + different salts and ions depending on the local environment. For a billion years it was this enormous biochemistry experiment. Life pretty much HAD to happen, somewhere in all that.

Depending of whether life is a fluke or normal outcome. Until we know how life on Earth began, we will have no idea for what to look for in the sky.

http://www.scientificamerican.com/article/a-new-physics-theo...

This theory could be quite important or wrong. Or both. But it gives very interesting definition of what life is.

Reply on news.ycombinator.com