Other civilizations in the Milky Way?

[Andrew D]

A new model computationally evaluates several explanations for the Fermi Paradox.

http://www.economist.com/blogs/babba...l-intelligence

[Rhaedas]

Even if every planet had intelligent life on it, each would have its own path of development. A real Drake's equation that took everything into consideration would be hugely complex, and we'd miss some other factors. And it would be as meaningless a number as what it gives now. It would also be heavily biased to what we expect to find. Until we actually start finding other life elsewhere, we only have one example of what's possible.

Nothing wrong with speculation, but that's all it is. We just don't have enough data to make actual calculations.

[NEOWatcher]

Nothing wrong with speculation, but that's all it is. We just don't have enough data to make actual calculations.

I agree.
My question is what this model has that other models haven't in the past. What refinements or new information does it bring?

I think that 10% speed of light travel is a bold assumption.

[JustAFriend]

And to talk of civiliztions lasting 50million years - even 250,000- is incredibly bold
when we have only had radio for a century and can't even get humans beyond low orbit.

I remember Sagan in one of his books putting us a .6 or .7 on the civilization scale,
so we're not even up to 1.0 yet. We certainly can't even deliberately send a probe
to another system, let alone build starships on a 500-year schedule, so it's highly
presumptuous to say what a galactic empire could or could not do.

[Andrew D]

I agree.
My question is what this model has that other models haven't in the past. What refinements or new information does it bring?

I think that 10% speed of light travel is a bold assumption.

The 10% in the article was an error. The figure used in the model was 1%. Drakes equation gives a (shaky) estimate for the number of other civilizations. The model characterizes how a civilization grows in three dimensional space based on parameters like resource consumption and a population cap among other factors. Also, using the model the authors define a density function that gives the probability of contact within a given time of civilizations as a function of their initial distance and the parameters described above. I can email interested parties (PM me) a preprint of the paper in pdf form, it is currently under review for publication.

[Andrew D]

And to talk of civiliztions lasting 50million years - even 250,000- is incredibly bold
when we have only had radio for a century and can't even get humans beyond low orbit.

I don't think it's bold at all. The civilizations in the model don't have to 'survive' in unity, that is, we're not interested in the probability that some species leaves its home and arrives here speaking the same language. We don't impose any restrictions on the colonies other than 'extant'.

...it's highly presumptuous to say what a galactic empire could or could not do.

Isn't that exactly what you're doing though, in saying they wont last?

[Van Rijn]

Nothing wrong with speculation, but that's all it is. We just don't have enough data to make actual calculations.

Exactly. At this point it's a matter of throwing in whatever assumptions you like and going from there.

[Andrew D]

Exactly. At this point it's a matter of throwing in whatever assumptions you like and going from there.

Isn't that what we call science? Speculation guided by natural principles? Then evaluation and "respeculation".

[Van Rijn]

I don't think it's bold at all. The civilizations in the model don't have to 'survive' in unity, that is, we're not interested in the probability that some species leaves its home and arrives here speaking the same language. We don't impose any restrictions on the colonies other than 'extant'.

But the only civilizations we know about have life-spans of centuries, maybe millenia, if you assume optimistic definitions, and they're all strictly Earth based. Why are these civilizations expanding to other stars in the first place? What's their motivation? How much does it cost? Unless star travel is pretty cheap I would wonder why, if they can build a starship, don't they just build another habitat in their home system?

[Van Rijn]

Isn't that what we call science? Speculation guided by natural principles? Then evaluation and "respeculation".

How do you evaluate it?

ETA: I mean how do unambiguously evaluate speculation in this area? I don't see how you can do much more than speculate.

[Andrew D]

But the only civilizations we know about have life-spans of centuries, maybe millenia, if you assume optimistic definitions, and they're all strictly Earth based.

What do you mean by "assume optimistic definitions?"

Why are these civilizations expanding to other stars in the first place? What's their motivation? How much does it cost? Unless star travel is pretty cheap I would wonder why, if they can build a starship, don't they just build another habitat in their home system?

Why did we go to the moon? Why do we send probes into space? Why did Galileo drop things? Why do people buy expensive condos close to the beach when there are perfectly affordable homes inland? Why are you debating a paper you haven't read? These questions are of the same vain.

How do you evaluate it?

Are we speaking specifically about the model in question or the scientific process?

[Rhaedas]

Why did we go to the moon? Why do we send probes into space? Why did Galileo drop things? Why do people buy expensive condos close to the beach when there are perfectly affordable homes inland?

They are human things to do. We can't necessarily assume that other civilizations, if they can even get to the points of scientific curiosity or exploration, will follow though.

Are we speaking specifically about the model in question or the scientific process?

His point was that how can you test a speculation of a particular variable in a Drake type equation when you have yet to collect any data to test. We will (hopefully)...we're a lot farther than we were even a few years ago on the planetary side, but that's only the beginning of a lot of parameters.

[Andrew D]

They are human things to do. We can't necessarily assume that other civilizations, if they can even get to the points of scientific curiosity or exploration, will follow though.

Sure we can. We can then make predictions about how such a civilization might behave, we just have to acknowledge that the predictions we make are within the context of our assumptions (If you ready the last sentence of the article, it says: "Either that or the neighbours are a particularly timid bunch.") The paper doesn't actually contain any conclusions or feign any hypothesis about the existence of such civilizations, it evaluates some hypothetical solutions to the Fermi paradox which have been presented (in scholarly literature) with similar assumptions, and points out shortcomings in previous models of spatial emigration (mainly superfluous assumptions).

His point was that how can you test a speculation of a particular variable in a Drake type equation when you have yet to collect any data to test. We will (hopefully)...we're a lot farther than we were even a few years ago on the planetary side, but that's only the beginning of a lot of parameters.

The paper doesn't even mention the drake equation and offers no suggestion of how many extant civilizations there are. In no way does it evaluate, modify, or supersede the drake equation.

If you'd like to read what it does say, I'm happy to send you a copy, since you haven't read it. (And I'm the overspeculative one...)

[astromark]

You can be a man of science and a optimist..

I see no immediate danger to humanity including from ourselves..

I can express optimistic advancement in all things... and do. My name is Mark Lee.

[Don J]

And to talk of civiliztions lasting 50million years - even 250,000- is incredibly bold
when we have only had radio for a century and can't even get humans beyond low orbit.

That is true for a civilization based on money .Imagine the exponential advancement of science in a civilization not based on that limitative factor.

[Van Rijn]

Sure we can. We can then make predictions about how such a civilization might behave, we just have to acknowledge that the predictions we make are within the context of our assumptions (If you ready the last sentence of the article, it says: "Either that or the neighbours are a particularly timid bunch.")

Okay, then the article makes it clear that the idea an interstellar civilization must expand throughout the galaxy is purely an assumption? As is the idea that civilizations would continue for 50 million years?

[Van Rijn]

What do you mean by "assume optimistic definitions?"

Well, I've seen some definitions that count pretty much everything back to the earliest recorded history as part of the same civilization, while others might point to dramatic societal changes as indicating new civilizations: On one hand you might count a civilization as thousands of years old with one definition, or less then a century old with another definition. But however you do it, it's far short of the 50 million year number mentioned in the article.

[Van Rijn]

Are we speaking specifically about the model in question or the scientific process?

I'm speaking about the set of assumptions they're using.

The article says, "Using what they believe to be conservative assumptions (as low as one chance in four of embarking on a colonising mission in 1,000 years), they calculated that any galactic empire would have spread outwards from its home planet at about 0.25% of the speed of light. The result is that after 50m years it would extend over 130,000 light years, with zealous colonisers moving in a relatively uniform cloud and more reticent ones protruding from a central blob. "

I see no reason to make these particular assumptions, or how you could usefully evaluate, with present data, an argument based on these assumptions versus an argument based on very different assumptions.

[NEOWatcher]

Drakes equation gives a (shaky) estimate for the number of other civilizations.

I disagree. All the equation does is highlight the major factors that need to be taken into consideration when determing the result.

A lot of the items you mention are sub-equations of the civilization parameters, while density is part of the planetary parameters.
It's not the equation thats shaky, it's the assumptions leading up to it's parameters.

[Don J]

But the only civilizations we know about have life-spans of centuries, maybe millenia, if you assume optimistic definitions, and they're all strictly Earth based. Why are these civilizations expanding to other stars in the first place? What's their motivation? How much does it cost?

How much does it cost?

It cost nothing for a civilisation not based on money .

[NEOWatcher]

It cost nothing for a civilisation not based on money .

Cost is not exclusive to monetary issues. Cost in lives, resources, time, effort, etc.

[Andrew D]

Okay, then the article makes it clear that the idea an interstellar civilization must expand throughout the galaxy is purely an assumption? As is the idea that civilizations would continue for 50 million years?

No, it doesn't necessarily expand. There some conditions under which the modeled civ. stops expanding, it's explained in detail in the paper.

Well, I've seen some definitions that count pretty much everything back to the earliest recorded history as part of the same civilization, while others might point to dramatic societal changes as indicating new civilizations: On one hand you might count a civilization as thousands of years old with one definition, or less then a century old with another definition. But however you do it, it's far short of the 50 million year number mentioned in the article.

This is ridiculous. This paper is not in the field of astrosociology, so I really don't care about the extraterrestrial social strata, and I certainly don't make any assumptions regarding the social characteristics of a civilization. We are looking at the growth of a probabilistic random graph on a three dimensional lattice which conveniently models the spatial dispersion of a growing network, in this case a moving civilization or self replicating probes, among many other physical processes. It also does this much better than any previous model of probabilistic growth. I could easily build a longevity parameter into the model, if it would make you happy. But it's not like you'd read that paper anyway, since it appears blog articles are all you require to be satisfied with your understanding of a topic under discussion.

I'm speaking about the set of assumptions they're using.

The article says, "Using what they believe to be conservative assumptions (as low as one chance in four of embarking on a colonizing mission in 1,000 years), they calculated that any galactic empire would have spread outwards from its home planet at about 0.25% of the speed of light. The result is that after 50m years it would extend over 130,000 light years, with zealous colonizers moving in a relatively uniform cloud and more reticent ones protruding from a central blob. "

I see no reason to make these particular assumptions, or how you could usefully evaluate, with present data, an argument based on these assumptions.

We actually don't make these particular assumptions, I don't know where the author got that. I didn't write the article, sorry. The value he is describing is a parameter in the model, and can be changed to any of infinitely many values. The reason the results are important is that they show the spread of colonization is linear and independent of this particular parameter.

...A lot of the items you mention are sub-equations of the civilization parameters

What items are you referring to specifically? What do you mean by "sub-equations of the civilization parameters"?

Cost is not exclusive to monetary issues. Cost in lives, resources, time, effort, etc.

We include limiting factors such as resource consumption and time cost in the model.

Again, if you want to read the paper, PM me and ill send you a copy. (Inbox is yet empty...)

[IsaacKuo]

For what it's worth, I don't see any problem with this sort of scientific paper and modelling. So what if some of the hypotheticals are untrue? This is about making predictions based on hypotheticals, which in turn may help falsify or confirm those hypotheticals.

As for this particular paper and model, I'm not sure what is novel compared to previous growth models. Some factors that I wish were accounted for are stellar motion, vastly uneven star system value, vastly uneven stellar density, variations in desired star system resources, territorial factions, and warring factions.

Stellar motion is a big deal for models which purport to show dead ends or some ultimate limit to growth on gigayear scales.

Vastly uneven star system value relates to the fact that some star systems have many orders of magnitude more power or resources than others. Star systems with black holes or neutron stars offer cheap relativistic propulsion. Red dwarf systems are numerous and long lived. White dwarfs offer some interesting possibilities. Systems with warm jupiters or brown dwarfs offer inexpensive exploitation of oort cloud objects.

Vastly uneven stellar density relates to how star clusters offer large numbers of star systems within easy access, so they would be more promising places to expand to than elsewhere.

Variations in desired system resources relate to how self replicating probes may only be programmed to exploit particular resources, available only in some fraction of star systems or in interstellar space (interstellar space resources may be particular desirable for military probes, since they are stealthy). It may also relate to biological aliens only desiring planetary resources of a certain type.

Territorial factions relate to how a faction may not accept incursions of other factions into their general territory, even if the incursions are into unoccupied star systems. The above factors represent various reasons why a faction may leave gaps of unoccupied star systems. Add in territoriality, and these gaps may be purposefully enforced against incursions.

Warring factions are essentially similar to territorial factions, especially if there are three or more factions. With three or more factions, there's a strong incentive to maintain border neutral zones rather than actively fight rivals. Engaging in an active conflict tends to weaken both of the parties involved by draining them of fighting units and resources. This punishes the aggressor by making him vulnerable to the other parties. This gets into some complexities of game theory.

Those are the sorts of things I'd like to see in interstellar expansion models.

[Andrew D]

These are all great suggestions, the only problem is computability. The computation time already grows as the number of colonies, which is polynomial, but still huge.

This gets into some complexities of game theory.

This is really where I would like to go next. The problem is that computing a payoff matrix for a few civilizations with more than a few dozen colonies each is not computationally feasible. Were running the computations using matlab on PC's, so were very limited by processing and memory. A 250,000 year model takes about 15 hours to compute, and that's if we limit the density a good deal.

[kamaz]

I'm going to point out two problems with interstelalr expansion:

1. Supply chain. I don't know how the interstellar spaceship will work, but I do know that it will be made in some spaceshipyard, and that spaceshipyard will have suppliers, and these suppliers will have their suppliers, and so on -- because a spacecraft has a lot of different parts, and different parts require different expertise and machinery to fabricate. Problem is, you can't take all of this to another solar system. You will take some equipment, some people, and some basic knowledge -- enough to survive, but not enough to be able to build new spaceship once you get there. So once the colonists land, they will experience a technological regress. They will need between several hundrend to several thousand years to be able to build their own interstellar spacecraft. (Provided of course that they would be willing at all -- a completely new solar system will provide them with a lot of resources. It will take millenia before they start running out of space.) However, by the time they are capable of building an interstellar craft, the home world will have colonized all nearby targets by churning out mission after a mission, each one cheaper with the previous, and each one with more range than the previous (scale effect).

2. Communication. Samuel P. Huntington had a valid point by pointing out that humans currently have seven civilizations on the planet. These civilizations are generally separated by natural barriers (oceans, mountains, deserts). The reason is obvious: a civilization needs communication; if a group of people is split by some barrier into two groups, these groups will begin to diverge. Culture will diverge within decades, language within centuries, genetics within millenia. Without FTL, interstellar distances are an obstacle to communication -- sufficient to cause a colony civilization to diverge from the mother world. By the time the colony (re-)learns how to build interstellar crafts, it will be a wholly different culture, speaking a different language, and, to use an old SF cliche, their spacecrafts will look nothing like our spacecrafts. The colony population may even become genetically distinct, because living on other world will cause different selection pressures. Once they go to the neighboring world to meet their kin, they both will see each other as aliens.

[kamaz]

Were running the computations using matlab on PC's, so were very limited by processing and memory. A 250,000 year model takes about 15 hours to compute, and that's if we limit the density a good deal.

Ouch. Some hints:

1. If running out of memory, make sure you're not swapping. Swapping slows you down by a factor of 1000.
2. MATLAB interpreter is glacially slow when doing loops. Instead of looping, vectorize inner loops. If you can't, rewrite inner loops in C and compile as MEX files.
3. See if you can paralellize the problem to run on several cores at once.
4. If the problem is paralellizable, and you have money, consider using Amazon EC2. E.g. if your model takes 100 hours to run on a PC, use EC2 to simultanously run 100 virtual machine instances and get the results in one hour.

[Don J]

I'm going to point out two problems with interstelalr expansion:

1. Supply chain. I don't know how the interstellar spaceship will work, but I do know that it will be made in some spaceshipyard, and that spaceshipyard will have suppliers, and these suppliers will have their suppliers, and so on -- because a spacecraft has a lot of different parts, and different parts require different expertise and machinery to fabricate. Problem is, you can't take all of this to another solar system. You will take some equipment, some people, and some basic knowledge -- enough to survive, but not enough to be able to build new spaceship once you get there. So once the colonists land, they will experience a technological regress. They will need between several hundrend to several thousand years to be able to build their own interstellar spacecraft.

That problem can be easely solved.
If they send 100 spaceships to the targeted solar system that they want to colonize,then they dont need to build new spaceships and they are ready to expand to another solar system once the new colony is well established.

[Robert Tulip]

A new model computationally evaluates several explanations for the Fermi Paradox.

http://www.economist.com/blogs/babba...l-intelligence

Hi Roobydo - I read your paper and the Economist review. 0.25% of the speed of light, your estimate figure for civilizational expansion, is 2.7 million kilometers per hour. This is an extremely fast speed. Why don't you do the calculation with a more realistic speed, say 10,000 km/h?

Just say we stabilized our planet and managed to have a thousand years of peace so we could then start to look at expansion. If we then gradually colonised the solar system over the next million years, we might then be able to build an asteroid that could leave the solar system at such a speed. To travel one light year at 10,000 km/h would take over 100,000 years, so we could reach Sirius in less than a million years. To fill the galaxy at this speed would take longer than the life of the universe (15 billion years).

Assuming an ability to travel between stars at millions of kilometers per hour seems speculative. If only slower speeds are possible, the frontier of alien civilizations would leave far bigger holes.

[Andrew D]

Hi Roobydo - I read your paper and the Economist review. 0.25% of the speed of light, your estimate figure for civilizational expansion, is 2.7 million kilometers per hour. This is an extremely fast speed. Why don't you do the calculation with a more realistic speed, say 10,000 km/h?

Just say we stabilized our planet and managed to have a thousand years of peace so we could then start to look at expansion. If we then gradually colonised the solar system over the next million years, we might then be able to build an asteroid that could leave the solar system at such a speed. To travel one light year at 10,000 km/h would take over 100,000 years, so we could reach Sirius in less than a million years. To fill the galaxy at this speed would take longer than the life of the universe (15 billion years).

Assuming an ability to travel between stars at millions of kilometers per hour seems speculative. If only slower speeds are possible, the frontier of alien civilizations would leave far bigger holes.

Literature says that civilizations could be millions of years older than us. Think of our technological progress in just the last hundred years. Voyager is leaving us at 61000 km/h. The speeds may be optimistic but I don't think they're unrealistic.

ETA: The model shows that the size of the holes depends mostly on the limiting value and not on travel time. The voids are emergent from resource consumption or other limiting factors, not including travel time.

[Van Rijn]

Literature says that civilizations could be millions of years older than us.

What literature? Based on what evidence?

Think of our technological progress in just the last hundred years. Voyager is leaving us at 61000 km/h. The speeds may be optimistic but I don't think they're unrealistic.

But again, without knowing where technology will go, it's all speculation at this point.