Hacker News Reader

Comment by blauditore | Hacker News Reader

Why is it generally assumed that organic compounds and (liquid) water are a necessity for any form of life?

Of course, all life on earth is based on these things, but that doesn't necessarily mean that something considerable as a life form couldn't arise from complex chemical processes with different elements and at different temperatures.

For example, electronic circuits are non-organic and independent of water, but still capable of complex behavior. Something remotely similar to human-made robots using e.g. sun light as power source might have spawned on a planet with lots of metals and silicone. Of course this exact robots scenario is highly unlikely to occur - in reality their brains/circuits would probably work in a more evolution-friendly way - but you get the gist.

View on HN

The alternatives tend to have a lot of problems that seem very difficult to overcome: https://en.wikipedia.org/wiki/Hypothetical_types_of_biochemi...

It's not just some sort of failure of the imagination. The idea that life might be really weird isn't a new idea; everybody's read and/or seen the same science fiction you have too. But if it's going to exist in this real universe, it has to follow the real rules of the universe, and it's not particularly obvious that there are any alternatives that can even theoretically work.

Somewhere out there silicon based life is having the same argument about carbon based life.

Did you read the link I provided? Intelligent silicon life has good reason to believe that carbon life is quite a bit more likely than their form of life. Just as we have good reason to think intelligent silicon life is very unlikely.

My point is precisely that this isn't just random chauvinism, which is rather than being some sort of way-cool open-minded "whoa" opinion is really rather insulting to the huge number of scientists who have considered this matter very carefully. There are a looooooooot of good reasons to think carbon is the only possible substrate for life in this universe, and a lot (still italicized, but with fewer "o"s) of good reasons to think water is the most likely choice.

To be fair, the estimated reading time for the link you provide is roughly 20 minutes and contains roughly 5000 words.

If silicon life makes so little sense then maybe we could get further if we were using organic compounds for computational equipment. I.e. Maybe there's a link between computational ability and the factors (raw materials and 'above') that make the basic components.

There's some research going into this now.

Here's an article about it from this past summer: https://www.sciencedaily.com/releases/2016/07/160714120650.h...

A researcher at GA tech has built a simple computer out of neurons: https://en.wikipedia.org/wiki/Wetware_computer

And I remember reading about storing data in DNA strands, but I can't find the article at the moment.

Anyway, it's a really interesting area of research. I expect we'll only see more of this stuff in the near future.

Saying that silicon life is less likely than carbon based life and that general AI is inevitable in the same breath would be a contradiction. I know you didn't mention AI yourself but there are plenty here on HN that would espouse those views.

But if we produce general AI in the near future it will have taken us billions of years, a considerable fraction of all time in this universe.

That would be assuming that the evolutionary progression to organic life as intelligent as humans must follow a similar path. Is there evidence that evolution is directed and that intelligence is some sort of "higher form", or is intelligence an arbitrary adaptation like any other, and could have occurred earlier in history? I could be wrong but I believe the latter is the current interpretation of evolution, in that humans could eventually evolve to something with _less_ intelligence.

If you know anything about silicon chemistry, you'd know it's a poor substitute for carbon.

For example, CH4 (methane) is a nice stable molecule. It exists underground for millennia. SiH4 (silane) is an unstable, pyrophoric (spontaneously ignites in air). There is no evidence to suggest that a silicon-based life form could exist.

> SiH4 (silane) is an unstable, pyrophoric (spontaneously ignites in air)

I'm no chemist, and I buy the line of reasoning that Carbon is by far the most likely substrate for life, but this statement seems to assume an oxygen-rich atmosphere. That's another assumption that could, in theory I guess, be invalid.

That's a good point, but another thing to consider is what the pyrophoric nature of silane indicates - highly unstable molecule. Even in an oxygen-free atmosphere, silicon hydrides just aren't that stable.

Why would silane be such an important factor in silicon based life?

The point made is the silicon-hydrogen bonds are much weaker than carbon hydrogen bonds.

This seems to, at the very least, discount silicon as a drop-in replacement for carbon.

So do we expect AGIs to be built on architectures that require silicon hydrogen bonds?

I use to think the same as you for the same reason as you, but there's actually a surprisingly strong argument that life forms that develop independently should still look like us at the lowest level. The idea is this: Although it's likely possible for alternative forms of life to develop, we should expect the probability of abiogenesis (i.e., the chance that life develops in a given pool of chemicals, per unit time per unit volume) to vary dramatically depending on the form of life. This is because abiogenesis is (almost certainly) an extremely improbable event which requires individually unlikely random events to all occur in sequence.

In other words: given the fundamental particles and interactions of physics, it would be very surprising if the chance of abiogenesis for two different forms of life were within, say, an order of magnitude of each other, since there are so many conditional probabilities that need to multiply for a given sequence of improbable events. This implies that, looking across the universe, there is a preferred form of life that is much more likely to be generated than all other forms. Therefore, with high probability it will be the form of life that we are and that any aliens we meet are.

Hm, interesting approach. So the assumption is that abiogenesis is highly complex, thus probabilities for fundamentally different life forms (likely) vary wildly. Hence either foreign life form X were far more likely to develop and we would see it all over the place (actually, maybe not), or it were far less likely and hence we'd rather find one similar to ours.

But how does one estimate probabilities of abiogenesis, or their relative magnitudes? Are there e.g. scientific papers published on this subject?

> Hence either foreign life form X were far more likely to develop and we would see it all over the place (actually, maybe not), or it were far less likely and hence we'd rather find one similar to ours.

Yea. The argument is actually independent of whether the probability of abiogenesis is large enough for life to independently arise multiple times in the solar system or (conversely) small enough that it's very unlikely for it to exist elsewhere in the observable universe. I think all it requires is that abiogenesis is a combination of several small probabilities distributed over several orders of magnitude. (If the probability is large enough that multiple forms routinely develop on the same planet, then we probably also have to assume something about, say, the first-mover dominating.)

> But how does one estimate probabilities of abiogenesis, or their relative magnitudes?

It's really hard.

> Are there e.g. scientific papers published on this subject?

Yes, but I don't know them off the top of my head, or whether any are any good.

"But how does one estimate probabilities of abiogenesis, or their relative magnitudes?"

You probably covered some of this in school, though it tends to be an aside in your chemistry or biology book and easy to miss. But people have done the basic experiments with taking a bunch of likely looking chemicals, based on what we think the old Earth looked like, and shooting electricity through them and other such things that seem like they may have contributed to life.

Those experiments came up negative; life isn't that easy. That has been so thoroughly absorbed into modern consciousness that it can be easy to forget that they were perfectly legitimate experiments to conduct, and had they come out positive, that pretty much would have ended all debated about how common life is in the universe, moving the discussion firmly into the "intelligent" part of the issue.

You can also try to create "minimal organisms" that are sufficiently simple that they perhaps could be randomly created.

However, last I knew, computing odds of abiogenesis is still effectively impossible because there are still some steps that we simply do not know how to get past via random processes, such as ensuring the chirality of the molecules matches. That's not to say there's a lot of ideas out there on how to get over various elements of the problem, but at the moment I think it's fair to say that the best science we current have puts the odds of abiogenesis at simply "impossibly unlikely". Stay tuned for more "best science", of course.

> effectively impossible because there are still some steps that we simply do not know how to get past via random processes

To be clear, there is always a lower bound on this probability that comes from brute force randomness (e.g., a quantum fluctuation in which all the molecules of a bacteria assemble spontaneously). This lower bound is extremely, extremely tiny, taking the form 10^(-10^(large)), and there are almost certainly easier ways to make a simpler replicator that later evolves to become a bacteria. But since we don't have an upper bound on the volume of the universe, and we don't have evidence for life arising somewhere else in our observable universe, it's formally possible that it's simply that stupendously rare (but still not impossible).

This doesn't prove anything, but it's interesting to consider at least: if you look at the list of the most common elements in the Milky Way[1], 7 of the top 10 are essential for life on Earth: Hydrogen (#1), Oxygen (#3), Carbon (#4), Iron (#6), Nitrogen (#7), Magnesium (#9), and Sulfur (#10). Silicon (#8) is also pretty much essential. The other 2 elements in the top 10, Helium (#2) and Neon (#5), are noble gases with low chemical reactivity.

So, what could that mean?

If you claim that those elements (or at least a subset) are essential for any kind of life, then by the anthropic principle, it's not surprising to see the distribution that we see in our home galaxy.

Alternatively, you could claim that life can evolve with all sorts of combinations of elements, but that it makes sense that it would evolve to use the most common elements in any given area of the Universe.

Either way, it seems like the fact that the life we know uses the most common elements in the galaxy is a good reason to expect non-Earth-based Milky Way life to be similar.

This is just a thought I've had in the past; I haven't done a whole ton of research into this correlation. And of course, there's a whole other set of arguments involving the chemical properties of elements, which I'm just skipping over in this comment.

[1] https://en.wikipedia.org/wiki/Abundance_of_the_chemical_elem...

EDIT FOR CLARITY: The sentence mentioning the anthropic principle used to read: "By the anthropic principle, you might say that those elements (or at least a subset) are essential for any kind of life, and given that life exists on Earth, it's not surprising to see the distribution that we see in our home galaxy."

I think the anthropic principle should only be applied the other way around: We're here due to this environment, so other places with similar environments might harbor life too. But this doesn't say anything about whether different environments may enable life too.

The second claim seems much more sense to me (like horses vs. zebras). However, actively searching for organic compounds in search of life would lead to circular reasoning.

After all, organic life is probably the easiest one to search for because it's the only one we know so far.

You're right about the anthropic principle; I wasn't clear. I've edited that sentence to clarify what I meant. As for none of this saying "anything about whether different environments may enable life", I agree; as I said at the beginning of the comment, this doesn't prove anything. Anyway, I was more talking about the likelihood of similar-to-Earth life as opposed to whether other types of life are possible at all.

Because only carbon chemistry allows for the creation of complex molecules at moderate energy levels.

Maybe a silicon based life-form is possible, when temperatures are around 1000K or even higher, for example.

That's the direction I'm thinking at too. For instance, even Venus has temperatures around 500 K at its surface, and much higher pressure than Earth. But it obviously gets harder to model chemistry the more different environments are from what we have here, due to the lack of (or difficulty to produce) empirical data.

It boils down to reducing your search space. By definition you can't search for something you know nothing about. (Another form of life than the one we know)

So when we search for "life" you really need to read "life as we know it"

A new form of life would be a massive discovery, but is by nature almost impossible to target as to a of the research.

Totally correct, but we haven't seen (or recognized) non-carbon based life yet, so it's reasonable to have a bias towards something we already know for a fact as opposed to theoretical predictions that have yet to be realized. I absolutely believe in the possibility of electronic sentience, but we'd be foolish to ignore all the biological evidence that surrounds us, whose chemical signatures we know how to identity.

because both water and carbon have very specific, near-unique properties at both the molecular and macro level that make them unique candidates for forming the basis of life.

Silicon and ammonia share some of these properties, so they potentially are competitors, but what we observe not just on earth, but as gas in interstellar space, is that complex molecules are all carbon-based.

Are earth computers not a product of evolution? It could be hypothesized that silicon-based or other "artificial" form of life is the next generation and organic life is only a temporary step in evolution.

You are right. But these sort of arguments go into science fiction and philosophical domains rather too quickly.

For eg: Life on neutron stars.

There for you have to settle for a search space and look there.

Reply on news.ycombinator.com