Thursday, October 16, 2008

The Living Dead

Meet Candidatus Desulforudis audaxviator, the bacterium that does it all: fix carbon, fix nitrogen, synthesize all essential amino acids, locomote — an organism that can exist totally independent of other life. It doesn't even need the sun. This fucker basically lives on sulfur, rock, and electrons*.

It's an obligate anaerobe, without even the most rudimentary oxygen resistance. A bug like βehemoth would kick its ass throughout most of the terrestrial biosphere (its natural digs are a couple of kilometers down in the crust, where no O2 has poked its corrosive little head for at least three million years). But that's not likely to be any kind of drawback out in space, and various talking heads are already nattering excitedly about the prospect of something just like this hanging out on Mars, or on the Saturnian moons.

It is cool. It is, quite literally, a complete ecosystem bundled into a single species, a biosphere crammed into two-and-a-half megabytes and a crunchy shell. Astrobiologists the world over have been creaming their genes for a week now. It's such a science-fictional little beast that its very name was lifted from a Jules Verne novel— but what really sticks in my mind about this little Swiss-army knife is a feature that's actually pretty common down there.

If it's anything like other deep-rock dwellers, D. audaxviator reproduces very slowly, taking centuries or even millennia to double in numbers. It's a consequence of nutrient limitation, but might we be looking at a kind of incipient immortality here? The textbooks tell us that one of the defining characteristics of life is reproduction. But if you think of life as the propagation of organized information into the future — the persistence of signal, rather than merely its proliferation — then reproduction is really just a workaround. The chassis that carries the information wears out, and must be replaced.

It doesn't take much, here at the dawn of Synthetic Biology, to imagine an organism with unlimited self-repair capabilities; something that can keep its telomeres nice and long, which sweeps away all those nasty free radicals and picks up the broken bottles in their wake, which replaces an endless succession of disposable Swatches with a solid gold Rolex which can hang in there for a billion years or more. Hell, you could even postulate some kind of Lamarckian autoedit option on the genes, so the organism can adapt to new environments. Or you could just limit your organism to extremely stable environments that don't require ongoing adaptation. Interstellar space, for example. Or deep in a planetary lithosphere. In some ways, this could be a superior strategy to conventional breeding; at least you wouldn't have to worry about population explosions.

I wonder if, somewhere down there, D. audaxviator or something like it has given up on reproduction entirely. Maybe it keeps the machinery around as a kind of legacy app that no one uses any more and just ticks slowly onwards, buried beneath all that insulating and protective rock, unto the very end of the planet.

The textbooks would call it dead. I'd suggest our definitions may need an upgrade.


*Of course, the fact that it can live independently doesn't mean that it evolved independently. A bunch of its genes have been cadged from Archae via lateral transfer. Its genes also contain anti-viral countermeasures; whether it siphoned those off incidentally from donor species or actually uses them to guard against parasitic code, there's obviously a history of contact with other life in this bug's family tree.

Labels: , , ,

19 Comments:

Anonymous Anonymous said...

"...creaming their genes..."

Oh, you.

Also, I like the idea of a Voyager species, but if it comes across a new predator isn't it toast? (That's my impression, at least.) And with slow reproduction aren't we facing potential species-wide genocide the moment it happens?

October 17, 2008 at 12:54 AM  
Blogger Matt McCormick said...

Cool.

So on the typical def. out of a college bio class, life has these properties:

Homeostasis
Organization
Metabolism
Growth
Adaptation
Response to stimuli
Reproduction

So you'd suggest that we scrap reproduction. Anything else?

MM

October 17, 2008 at 12:57 AM  
Anonymous Anonymous said...

Mr. McCormick asks:

So you'd suggest that we scrap reproduction. Anything else?

Wait a moment and our host will propose, I am thinking,
that life is "information shaped by natural selection"?

Then we can list the items that under that definition that might be alive -

virus
prion
fire (I know you said, Mr. Watts, but I think it fits)
culture
threads in a blog posting

October 17, 2008 at 9:49 AM  
Blogger Peter Watts said...

OK, so I'm just finishing off my response to the first couple of comments here and bec-87rb jumps up and steals all my thunder. Well, she's right about the definition I'm gonna invoke; she's wrong about fire fitting that new definition. She may be wrong about prions/viruses, depending on where you want to draw the physical boundaries around the entity you wish to call "living". Personally, I'd be inclined to let viruses into the club at least, even if that means defining Britney Spears pop singles as a kind of viral self-replicating (parasitic) life-form.

But to maintain the integrity of my original post, I offer it here, unedited even in light of becster's impertinent jumping of the gun:


Madeline Ashby
said...

Also, I like the idea of a Voyager species, but if it comes across a new predator isn't it toast?

Depends on how easily the predator can eat it. It doesn't depend on a surrounding ecosystem, so its biochemistry might well be utterly incompatible with any predator from said ecosystem. Predators might not even recognize it as food. Also, keep in mind that this bug is a strict anaerobe, and conventional wisdom holds that anaerobic metabolism puts huge constraints on the development of complex life. Any predators likely to overlap this thing's niche will probably be pretty rudimentary.

And with slow reproduction aren't we facing potential species-wide genocide the moment it happens?

We're not. The immortal nonreproducer might though, yes; a breeding competitor could swamp it through sheer numbers, even if it wasn't as efficient on a head-to-head basis. But there does seem to be a growing realization that "fill the world up with thy numbers" is a pretty crude interpretation of Darwinian imperatives (Microsoft, for example, managed to turn the PC nichescape into a monoculture and it's still digging out from all the crashes and virii that attracted). Things always look rosiest for a bacterial culture just after it's gobbled up half its nutrients; then, next reproductive iteration, it's out of food and the population crashes. In contrast, and counterintuitively, when there's a very small number of something they can be extremely hard to kill off; there's always some little refugium somewhere, some little hiding spot behind the fridge, where it can wait out whatever apocalypse takes out the competition. And if your competition reproduces sexually you're even better positioned; you can drive a sexually-reproducing species to extinction simply by reducing the population level enough to make it difficult for the survivors to find mates. You can take out an asexually-reproducing species by reducing nutrient levels to the point where they can't replicate. It takes less energy for a body to just keep idling than it does to build a copy of itself, so an immortal nonreproducer would be able to outlast certain stresses that would take out breeders of any stripe.

Of course, it all depends on the breaks of any given scenario. But reproductive success at t=1 doesn't do you much good if you've eaten yourself to extinction at t=2, so perhaps the real criterion for success shouldn't be "Am I outbreeding the competition?", but merely, "Am I dead yet?" And in that context, D. audaxviator might prove more successful than all of us.


Matt McCormick said...

…So you'd suggest that we scrap reproduction. Anything else?

Actually, I'd suggest we scrap the whole checklist approach; a list of attributes, after all, doesn't constitute a definition, and the problems with such lists is that they tend to get unwieldy when you encounter an exception, and have to add another item to account for it (ooh, look at this; a salt crystal "grows" by "feeding" on ions from its environment, and can repair "damage" to its structure; fire "grows", "reproduces" by budding, "metabolizes" food and oxygen into CO2* and waste heat— better add, oh, I dunno, homeostasis to weed them out — and maybe we ought to make reproduction optional, or we'd have to revise the criminal code to reflect the fact that people with vasectomies can't be killed, since they're not alive to begin with…)

You don't define life by listing the things it's got (or doesn’t have); you define it by describing what it is. And the best definition I've encountered was put forth by Dawkins a couple of decades ago: Life is information shaped by natural selection.

Of course, some would object that this would force us to describe computer viruses as literal life forms. Not necessarily. "Natural selection" implies more than reproduction; it implies heritable copy errors, selection filters, and so on. A computer virus is not alive if someone has to keep rewriting the fucker to account for new countermeasures; it is alive, I would argue, if its lineage evolves over time without explicit human intervention. So, yes; in principle, I have no trouble describing electronic entities as literal life forms.

I'm not even being particularly radical in this approach. For years now, top-of-the-line scientific journals have been publishing papers which explicitly treat A-life as a legitimate subset of the real thing.


(*Would you believe that fucking blogger won't accept the <'sub'> tag in comments? I have to add quotes to the tag even here, or it won't let me post! How lame is that? With each passing day, blogger sucks harder and harder on the hairy banana.

I've been hearing good things about Wordpress; anybody have any experience with that?)

October 17, 2008 at 10:18 AM  
Blogger Unknown said...

Dumb Question:
Can gene/DNA data be easily converted to binary code now for any species or is it still a laborious process? Is anyone doing that for species other than humans?

October 17, 2008 at 10:43 AM  
Anonymous Anonymous said...

I agree with scrapping the use of lists of criteria to establish whether something exists or not. For example, years ago someone (I forget who) argued that to be inteligent, a species must be self aware, use language and produce tools (and probably other things that I can't remember right now). However, when we found animals that met these criteria, we simply changed them.

Like the argument over who is smarter, cats or dogs. Dog lovers argue that dosgs are smarter because they can be trained. I argue that cats are smarter because they have figured out that they can get what they want without sitting on command.

October 17, 2008 at 10:46 AM  
Blogger Keippernicus said...

Oooh, ooh, me me me! Wordpress is totally awesome, robust and customizable. I don't know if it'd be six and a half kinds of ass pain to port the crawl but once there its a snap.

I do like the idea of an immortal species that can update its own genetics and auto-adapt to a changing environment.

However for this hypothetical species to truly persist there has to be some means for it to jump ship when its patron star decides to fizzle or blow up (or if it gets iron-bombed by Charlie Stross' Eschaton.) For practical purposes I would suggest an inorganic alternative.

For as we are at the dawn of synthetic biology we are also nearing the dawn (hopefully) of purely inorganic life as well.

I may be wrong but the idea of a microscopic anerobe drifting calmly through space doesn't seem quite as competitive as say, a nugget of photovoltaic computronium that can survive atmospheric reentry.

After all, no matter how deep you go in a planetary crust it will not matter once some species comes along that can eat the entire planet and turn it into another huge mass of intelligent robotic goodness.

Then again, what good is a planet eating form of life when compared to a star-eating species? Maybe I'm not playing by the biological rules but it gives me chills to think that somewhere out in the void there could already be intelligent machines chewing up distant galaxies and spreading across the universe.

It's comforting to know that they probably won't get here until after the sun's fried out the planet but if computer viruses have taught us anything about persistence and replication it seems the first species or technology capable of freely reproducing throughout interstellar space will do so.

October 17, 2008 at 11:24 AM  
Anonymous Anonymous said...

Mr. Watts:

OK, so I'm just finishing off my response to the first couple of comments here and bec-87rb jumps up and steals all my thunder

A man's blog is his castle, his head space, and we are your guests, ergo, your thunder is always preeminently loud here, my friend.

And hey, *smile* if you were just a little faster, man...

October 17, 2008 at 11:48 AM  
Anonymous Anonymous said...

Also, please defend your claim that fire is not information and that it is not shaped as it burns by its environment.

Also, please defend the idea artificial life is being shaped by natural selection. If I turn my electronic life loose in an electronic environment and it is eaten or eats other electronic life, how does nature have anything to do with that selection.

You could start by, mmm, defining "information"? Your choice.

Did I mention I share your enthusiasm for the idea of an anaerobe that can live in space - that *is* exciting.

Last, you get +2 thunder points for knowing that criterion is the singular of criteria.

October 17, 2008 at 11:58 AM  
Anonymous Anonymous said...

I say viruses count. Not least because it means that the MS monoculture is actually supportive of infodiversity by having so many gaps and holes for them to live in. (It's a shipwreck!) Clearly this means that censorware and copyright lawyers are actually good catalysts for innovation, in that they provide inhospitable environments that require the creation of thirdspaces. (I have officially deployed my thesis!goggles. Congratulations.)

But I'd also say viruses because our little friend has interacted with them in the past and changed as a result. To me that means they constitute some threat. If the little guy is named for a Verne novel, I'm worried about a Wells scenario, where someone coughs in space (or steps on something with a non-sterile boot, or, or...) and eradicates them all.

That does mean that I'd be less worried about the Eschaton, though. It sounds as though as long as the niche itself isn't contaminated, the little guys are fine. So even if Festival or the Big E show up, the worst our friends would have to worry about are over-optimistic transhumanists (whose gear they could maybe grab onto and migrate elsewhere and eliminate that species-wide genocide problem, no?). And the thing about those is the limitation of imagination: even if you could ask for the mods to live in this particular niche, you would have to know about it first, and provide specific criteria for survival.

In other words it's gonna be you down there, sending us "I CAN HAZ CALCIUM, PLZ?" macros. At least until more of those transhumanists access this post.

October 17, 2008 at 12:05 PM  
Blogger Ohnemus said...

re: bec

so lets start with information. The definition of information is mathematical and subtle but it boils down to: "the information content of a system is proportional to the smallest representation of the system"

fire is not information nor is it a live because it isn't a system, it's a chemical reaction taking place IN a system. it can shape a system, but in the end fire is a verb not a noun.

as for e-life, yes a computer isn't what we normally think of as "nature" but thats not what natural selection means. the word natural in there was just shoved in to differentiate between artificial selection- things like directed breeding (or in computer land between the actual coding process)- and "natural" selection.

October 17, 2008 at 2:31 PM  
Anonymous Anonymous said...

ohnemus:

The definition of information boils down to: "the information content of a system is proportional to the smallest representation of the system"

Say more, please. Can you unboil some? I'll stop you when I don't get it.

October 17, 2008 at 3:26 PM  
Anonymous Anonymous said...

bec-87rb said:
Also, please defend the idea artificial life is being shaped by natural selection. If I turn my electronic life loose in an electronic environment and it is eaten or eats other electronic life, how does nature have anything to do with that selection.

One could reasonably argue that 'natural' selection has nothing to do with 'Nature' as we generally think of it. Instead, the process of 'natural selection' could more easily be considered the interplay between one's environment and the ability of the organism to survive in that environment. Currently, natural selection affects biological entities through their reproductive likelihood, as self-modifying immortals haven't yet been identified, nor have non-biological organisms. The process should be applicable to any 'life', no matter what environment it arises in.

One thing to consider about the original idea - how did the organism initially evolve to the state of immortality? It would have to have developed from a pre-existing organism that did undergo reproduction (otherwise we open up the whole 'spontaneous generation' can of worms). Presumably such precursors would likely still be present in the environment of our now-immortal organism. The precursors may form the pool from which a successful predator of the immortal offshoot would emerge.

Another item - if the immortal form did evolve from a reproducing mortal form, then it is likely to still be capable (and therefore likely) to reproduce, unless the mutation that led to immortality is linked to a loss of reproductive capability. Thus, you have the situation of a reproducing immortal - not a situation that is good for the surrounding biota. They will be under extreme selective pressure to develop a way to predate the immortals (thus negating the 'immortal' bit) or prevent their numbers from increasing. One other bit along similar lines - organisms will expand their numbers until they reach a block to expansion (predator, lack of an essential component to growth or reproduction, etc.). I'm not aware of any self-limiting organisms. Presumably your immortal form would be subject to similar drives and inhibitions, although for a space-going organism the time-scale might be significantly different than the time-scale that we consider in our planet-bound environments.

October 17, 2008 at 6:14 PM  
Anonymous Anonymous said...

I'm not sure viruses should count but I think Mimiviridae should serve as dividing line. Its genome size and capability rivals some bacteria and almost doesn't need a host.

Having said that, wouldn't classifying "life" as complex semi self contained viruses make more sense or is that too scramblerish?

October 18, 2008 at 1:00 AM  
Anonymous Anonymous said...

davidk replies:

'natural selection' could more easily be considered the interplay between one's environment and the ability of the organism to survive in that environment. Currently, natural selection affects biological entities through their reproductive likelihood, as self-modifying immortals haven't yet been identified, nor have non-biological organisms. The process should be applicable to any 'life', no matter what environment it arises in.


Okay. So it sounds as if this definition further hinges on the definition of "organism." I am still trying to get a handle on why we have to eliminate fire, under the definition.

I can see in a global, gut sort of way, why fire can't be said to undergo the process of natural selection, but not in a logical sense, unless we define the point-of-action (selection) in terms of outcome (evolution, or material change in the item as the result of selection pressure.)

October 18, 2008 at 10:38 AM  
Anonymous Anonymous said...

I forget who mentioned it first, but I second the layout issue. I was looking for the "pedophilia in a pill" entry for something at my own place, and my vision started to diverge.

October 18, 2008 at 11:20 PM  
Anonymous Anonymous said...

"Britney Spears pop singles as a kind of viral self-replicating (parasitic) life-form."

And you said that you were not a musical snob. OK. I concede the point that whatever it is that Spears does really can't be classified as music.

October 20, 2008 at 2:39 AM  
Blogger onegaishimasu said...

Can I double-quote? Let's see if this works...


OpenID bec-87rb said...

ohnemus:

The definition of information boils down to: "the information content of a system is proportional to the smallest representation of the system"

Say more, please. Can you unboil some? I'll stop you when I don't get it.


---

Okay, so suppose I say 'one' 1500 times. That's not really 1500 bits, is it? Because I could say 'write one 1500 times' and (depending on what language I write that instruction in) that sentence takes much less than 1500 bits.

It's a problem of data compression. When I was new at computers, I noticed that a zip file was way smaller than the txt file stored within. So I figured 'why not zip the file twice?' I did. The file size actually increased a bit, the second time I tried compressing it. I didn't know at the time, but any compression algorithm can get a certain file down to a certain size and never any smaller.

You can define the minimum size that a bitstring can be compressed to as the 'information entropy' of that bitstring, and that (in a nutshell) is what I think ohnemus is talking about.

There's only really two subtleties to worry about. Different compression algorithms can compress the same file to different minima. Also, there's a certain amount of information specified within the algorithm itself.

As the joke goes, I have a video compression algorithm that can squeeze a movie down to one bit long. The movie is 'Hunt for Red October', and the algorithm takes 5 gigabytes to describe.

ohnemus, if I messed it up, correct me?

October 24, 2008 at 11:44 AM  
Anonymous Anonymous said...

Okay, so suppose I say 'one' 1500 times...[snip]video compression algorithm that can squeeze a movie down to one bit long. The movie is 'Hunt for Red October', and the algorithm takes 5 gigabytes to describe.

That make sense to me, anyway. So how does that relate to fire? Data about fire are too hard to compress, are beyond a certain compression threshold, hence, not alive?

October 30, 2008 at 3:12 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home