https://mason.gmu.edu/~rhanson/greatfilter.html

              The Great Filter - Are We Almost Past It?

                           Sept. 15, 1998

                           by Robin Hanson

    Humanity seems to have a bright future, i.e., a non-trivial
    chance of expanding to fill the universe with lasting life. But
    the fact that space near us seems dead now tells us that any
    given piece of dead matter faces an astronomically low chance of
    begating such a future. There thus exists a great filter between
    death and expanding lasting life, and humanity faces the ominous
    question: how far along this filter are we?

    Combining standard stories of biologists, astronomers,
    physicists, and social scientists would lead us to expect a much
    smaller filter than we observe. Thus one of these stories must be
    wrong. To find out who is wrong, and to inform our choices, we
    should study and reconsider all these areas. For example, we
    should seek evidence of extraterrestrials, such as via signals,
    fossils, or astronomy. But contrary to common expectations,
    evidence of extraterrestrials is likely bad (though valuable)
    news. The easier it was for life to evolve to our stage, the
    bleaker our future chances probably are.

Introduction

Fermi, Dyson, Hart, Tipler, and others [Finney & Jones, Dyson 66,
Hart 75, Tipler 80] have highlighted the relevance to SETI (the
search for extraterrestrial intelligence) of the "The Great Silence"
[Brin 83] (also known as the Fermi paradox), the fact that
extraterrestrials haven't substantially colonized Earth yet. What has
not yet been sufficiently highlighted or adequately analyzed,
however, is the relevance of this fact for much bigger choices we now
make.

The Great Silence must force us to revise a standard view in one or
more area of biology, astronomy, physics, or the social sciences. And
some of these revisions strongly suggest that humanity be much more
wary of possible disasters. To clarify these points, this paper will
first review how our standard understandings in these areas would
lead us not to expect a Great Silence, and will then consider a
variety of possible revisions we might consider.

Life Will Colonize

So far, life on earth seems to have adapted its technology to fill
every ecological niche it could. Previously stable populations and
species have consistently expanded into newly-opened frontiers. All
known life seems to have a "dispersal phase" to encourage
colonization, with non-trivial mutations and sexual mixing to
encourage exploration of new technologies [Tipler 80].

Similarly, humanity has continued to advance technologically, and to
fill new geographic and economic niches as they become
technologically feasible. For example, while imperial China closed
itself to exploration for a time, other competing peoples, such as in
Europe, eventually filled the gap.

This phenomena is easily understood from an evolutionary perspective.
In general, it only takes a few individuals of one species to try to
fill an ecological niche, even if all other life is uninterested. And
mutations that encourage such trials can be richly rewarded.
Similarly, we expect internally-competitive populations of our
surviving descendants to continue to advance technologically, and to
fill new niches as they become technologically and economically
feasible.

Colonization has been a consistent experience with life on Earth over
the long run, and our best understanding of human social systems
suggests this will continue. While humans evolve within complex
co-evolving organizational, cultural, memetic, and genetic systems,
all of these systems show long-term tendencies to make use of
reproductively-useful resources.

Thus we should expect that, when such space travel is possible, some
of our descendants will try to colonize first the planets, then the
stars, and then other galaxies. And we should expect such expansion
even when most our descendants are content to navel-gaze, fear
competition from colonists [Benford 81], fear contact with aliens, or
want to preserve the universe in its natural state. At least we
should expect this as long as a society is internally-competitive
enough to allow many members to have and act on alternative views.
After all, even navel-gazing virtual reality addicts will likely want
more and more mass and energy (really negentroy) to build and run
better computers, and should want to spread out to mitigate local
disasters [Zuckerman 85]. A million years is a cosmologically short
period, yet it is much more than enough for historic population
growth rates (> .001%/yr.) to overwhelm fundamental physical limits
on the amount of computation possible within the observable universe
[Zaslavskii 96]. This remains true even using black holes for
negentropy and quantum computers for computation, each of which
squares the available resources relative to standard approaches. Thus
we have good reasons to expect unused resources to be colonized on
cosmological time scales, even if we find other civilizations to
communicate with or to "teleport to" [Scheffer 94].

Evolutionary theory even suggests [Hansson & Stuart 90] that
competitive pressures among colonists should encourage a maximum
feasible economic growth rate, as those who travel too slow, linger
too long, or choose not to replicate [Stephenson 79] become
outnumbered by others. Increasingly fast and high risk colonization
probes may be sent on increasingly long journeys, all for a chance at
being the first to colonize vast virgin territory.

Technically, such space colonization seems feasible, even if it is
well beyond our current abilities, since even now we can envision the
enabling technologies. Slow self-sufficient interstellar boats would
be nearly feasible now, if we were rich enough to construct them. And
fast less-than-kilogram-sized [Forward 85,87] self-reproducing
[Tipler 80] nanotech-based [Drexler 92b] space-traveling machine
intelligences (artificial or uploaded [Hanson 94]) seem possible
within a few centuries.

There are no obvious limits to spacecraft speed (other than
lightspeed), given sufficient resources. And with full
(nanotech-based) control over the atomic structure of matter [Drexler
92a], colonists should mainly be interested in the atoms and
negentroy they can extract from a colonization site [Dyson 66,79],
and the convenience of its location.

The Data Point

Within the next million years (at most) therefore, our descendants
seem to have a foreseeable (greater than one in a thousand) chance of
reaching an "explosive" point, where they expand outward at near the
speed of light to colonize our galaxy, and then the universe, easily
overpowering any less developed life in the way. FTL (faster than
light) travel would imply even faster expansion.

We expect such an explosion to fill most every available niche
containing usable mass or negentroy resources. And even if the most
valuable resources are between the stars or at galactic centers, we
expect some of our descendants to make use of most all the matter and
energy resources they can economically reach, including those in
"backwater" solar systems like ours and those near us.

Once an explosion goes beyond the scale where a single disaster, such
as a supernovae, could destroy it, to become a "lasting" explosion of
advanced life, it should only be stopped by meeting another explosion
of similarly-advanced life. After that, if disaster befalls some
long-established colony, others should soon return to try again.

Without FTL travel to mediate conformity, we would also not be
surprised by a great diversity among the different parts of an
explosion, and especially among different explosions [Hoerner 78]. We
would expect, for example, different cultures, languages, and body
form details. We expect much less diversity, however, regarding
choices which would put a civilization or entity at a strong
competitive reproductive disadvantage.

For example, while one can imagine predatory probes sent to search
and destroy other life [Brin 83], it is harder to understand why such
probes would not also aggressively colonize the systems they visited,
if such colonization were cheap. Aggressive colonization would give
them all the more probes to work with, and deny resources to
competitors. If this colonization effort could hide its origins from
those who might retaliate, what would they have to lose?

Similarly, while some groups might plausibly leave some places
"fallow" as information-generating "nature preserves" [Fogg 87], it
is much harder to imagine that most places would be so preserved.
There should be diminishing returns to such information, and groups
that use more of their resources should be at a competitive
advantage. And given the vastness of space, substantial resources
should be required to keep "poachers" from slipping in to colonize
such a preserve.

Finally, we expect advanced life to substantially disturb the places
it colonizes. Whenever natural systems are not ideally structured to
support colonists, we expect changes to be made. And unless ideal
structures always either closely mimic natural appearances or are
effectively invisible, we expect advanced life to make visible
changes.

For example, it only takes a small amount of nuclear waste dropped
into to visibly change its spectra [Whitmire & Wright 80.] And a
civilization might convert enough of a star's asteroids into orbiting
solar-energy collectors to collect a substantial fraction of this
star's output, thereby substantially changing the star's spectral,
temporal, and spatial appearances. Even more advanced colonists may
disassemble stars [Criswell 85] or enclose them in Dyson spheres well
within a million years of arrival. Galaxies may even be restructured
wholesale [Dyson 66].

If such advanced life had substantially colonized our planet, we
would know it by now. We would also know it if they had restructured
most of our solar system's asteroid belt (though much smaller
colonies could be hard to detect [Papagiannis 78]). And they
certainly haven't disassembled Jupiter or our sun. We should even
know it if they had aggressively colonized most of the nearby stars,
but left us as a "nature preserve".

Our planet and solar system, however, don't look substantially
colonized by advanced competitive life from the stars, and neither
does anything else we see. To the contrary, we have had great success
at explaining the behavior of our planet and solar system, nearby
stars, our galaxy, and even other galaxies, via simple "dead"
physical processes, rather than the complex purposeful processes of
advanced life. Given how similar our galaxy looks to nearby galaxies,
it would even be hard to see how our whole galaxy could be a "nature
preserve" among substantially-restructured galaxies.

These considerations strongly suggest that no civilization in our
past universe has reached such an "explosive" point, to become the
source of a light speed expansion of thorough colonization. (That is,
no civilization within the past light cone of a million years ago for
us; see Technical Appendix below). Much follows from this one
important data point [Hart 75, Tipler 80].

The Great Filter

Consider our best-guess evolutionary path to an explosion which leads
to visible colonization of most of the visible universe:

 1. The right star system (including organics)
 2. Reproductive something (e.g. RNA)
 3. Simple (prokaryotic) single-cell life
 4. Complex (archaeatic & eukaryotic) single-cell life
 5. Sexual reproduction
 6. Multi-cell life
 7. Tool-using animals with big brains
 8. Where we are now
 9. Colonization explosion

(This list of steps is not intended to be complete.) The Great
Silence implies that one or more of these steps are very improbable;
there is a "Great Filter" along the path between simple dead stuff
and explosive life. The vast vast majority of stuff that starts along
this path never makes it. In fact, so far nothing among the billion
trillion stars in our whole past universe has made it all the way
along this path. (There may of course be such explosions outside our
past light cone [Wesson 90].)

The fact that our universe seems basically dead suggests that it is
very very hard for advanced explosive lasting life to arise. And if
there are other radically different paths to expanding lasting life
[Shapiro & Feinberg 82], that only makes the problem worse, by
implying that the filter along our path must be even larger.

Someone's Story is Wrong

Biologists and others have been working hard for a long time to come
up with plausible explanations for each of the evolutionary steps
listed above, explanations which make each step seem not especially
improbable. Plausible models have been offered of how RNA evolved to
reproduce, how simple (prokaryotic) cells grew around it, how cells
became more complex (eukaryotes), how cells came together into
organisms, how brains and hands evolved from simple control
mechanisms, and how our brains and hands lead to tool use and
scenario generation, which led us to where we are today.

Together these plausible explanations have persuaded countless teams
to construct relatively high estimates of the probability that any
one planet will eventually produce intelligent life such as
ourselves, by estimating relatively low values for each filter term
in the famous "Drake Equation" [].

Similarly, technological "optimists" have taken standard economic
trends and our standard understanding of evolutionary processes to
argue the plausibility of the story I gave above, that our
descendants have a decent chance of colonizing our solar system and
then, with increasingly fast and reliable technologies of space
travel, colonizing other stars and galaxies. If so, our descendants
have a foreseeable chance of reaching such an explosive point within
a cosmologically short time (say a million years).

Of course many other folks don't consider this scenario particularly
"optimistic" - they prefer that our descendants choose a more stable
path, less likely to "disturb the universe". But I will continue to
use the word "optimistic" to describe this scenario, because even
fans of stability should be concerned about the implications of
humanity not living long enough or free enough to have even a one in
a million chance, for example, that any descendant of ours will
escape to colonize space. It would seem that any reasonably
non-pessimistic scenario would include a non-trivial chance that at
least some of our descendants will choose the explosive path over the
next million years.

While all of these stories are at least minimally plausible, our main
data point implies that at least one of these plausible stories is
wrong -- one or more of these steps is much more improbable than it
otherwise looks. If it is one of our past steps, such as the
development of single-cell life, then we shouldn't expect to see such
independently evolved life anywhere within billions of light years
from us. But if it is a step between here and a choice to explode
that is very improbable, we should fear for our future. At the very
least, our potential would have to be much less than it seems.
Optimism (as defined here) regarding our future is directly pitted
against optimism regarding the ease of previous evolutionary steps.
To the extent those successes were easy, our future failure to
explode is almost certain.

Note that this cause for concern has a different basis than the
simple statistical arguments of Gott [Gott 93] and Leslie [Leslie 96]
that all else equal we shouldn't expect many more future humans than
there have been past humans. While those arguments shouldn't be
ignored, their strength depends much more on the auxiliary
assumptions one makes about other relevant information. In contrast,
the conclusion that the Great Filter is very large is relatively
insensitive to other assumptions.

It Matters Who's Wrong

Rational optimism regarding our future, then, is only possible to the
extent we can find prior evolutionary steps which are plausibly more
improbable than they look. Conversely, without such findings we must
consider the possibility that we have yet to pass through a
substantial part of the Great Filter. If so, then our prospects are
bleak, but knowing this fact may at least help us improve our
chances.

For example, if our prospects are likely bleak we should search out
and take especially seriously any plausible scenarios, such as
nuclear war or ecological collapse, which might lead to our future
inability to explode across the universe. A long list of such
scenarios for concern can be found in [Leslie 96]. Our main data
point, the Great Silence, would be telling us that at least one of
these scenarios is much more probable than it otherwise looks.

With such a warning in hand, we might, for example, take extra care
to protect our ecosystems, perhaps even at substantial expense to our
economic growth rate. We might be even especially cautious regarding
the possibility of world-destroying physics experiments. And we might
place a much higher priority on projects like Biosphere 2, which may
allow some part of humanity to survive a great disaster.

To find out whether such sacrifice is called for, humanity would do
well to study this whole area much more carefully, considering all
plausible explanations of the Great Filter. To encourage such study,
the rest of this paper will attempt to review the current status of
our understanding, considering in turn various possibilities
regarding who might be wrong, and the various types of evidence which
might clarify the matter.

Reconsidering Biology

First, let us review and reconsider our biological expectations,
keeping an eye out for prior evolutionary steps which may be more
improbable than they look.

Many theoretical stories have been offered to make various prior
evolutionary steps seem relatively likely, at least over a long term.
Given the complexity of the subject matter, however, these stories
are understandably sketchy. Thus the simplist way such theories might
be wrong is by having ignored some important factors and details. As
a general rule, simple plausible models quite often fail to capture
the essence of complex phenomena.

It should also be noted that many biologists expect a large, not
small, filter between dead matter and intelligent tool-using life
like us. They have complained that astronomers who estimate Drake
equation terms do not know enough biology, and they note in
particular that substantial tool use such as we see in humans has
only evolved once, and may well be a very unlikely evolutionary
accident [Simpson 64, Mayr 85,95].

In any case, it turns out that the very idea that a significant
portion of the Great Filter might reside in our past evolutionary
steps has important implications which can aid us in evaluating this
hypothesis [Carter 83, Hanson 96].

First, let us distinguish between two different kinds of evolutionary
steps. Let a "discrete" evolutionary step be one which must succeed
within certain a short time period; failure then implies failure
forever after. For example, if a certain type of solar system is
required, then success here can only happen when the solar system
forms. In contrast, let a "trial and error" step be something like
search across a mostly flat fitness landscape, where failure today
does not much affect the chances for success tomorrow. The main Great
Filter implications are regarding trial and error type steps.

Consider a situation where a certain number of trial and error steps
must be completed in a certain order within a certain total time
window. That is, for each step there is some constant probability per
unit time of completing that step, given that the previous step has
been completed. If the probability of completing all the steps within
the time window is low, then it turns out that for the cases where
all the steps are in fact completed, the average time to complete
each "hard" step is unrelated to how hard that step is!

For example, say you have one hour to pick five locks by trial and
error, locks with 1,2,3,4, and 5 dials of ten numbers, so that the
expected time to pick each lock is .01,.1, 1, 10, and 100 hours
respectively. Then just looking at those rare cases when you do pick
all five locks in the hour, the average time to pick the first two
locks would be .0096 and .075 hours respectively, close to the usual
expected times of .01 and .1 hours. The average time to pick the
third lock, however, would be .20 hours, and the average time for the
other two locks, and the average time left over at the end, would be
.24 hours. That is, conditional on success, all the hard steps, no
matter how hard, take about the same time, while easy steps take
about their usual time (see Technical Appendix). And all these step
durations (and the time left over) are roughly exponentially
distributed (with standard deviation at least 76% of the mean).
(Models where the window closing is also random give similar
results.)

We can apply this model to the evolution of life on Earth, by
examining the fossil record for roughly equally spaced apparent major
innovations. Such an analysis can complement other attempts to find
hard steps by intrinsic difficulty, necessity, and uniqueness in
evolutionary history, such as attempted in [Barrow & Tipler 86]

The fossil record shows about five roughly-equal periods between
major evolutionary changes since the Earth was formed [Schopf 92,
Skelton 93]. Specifically, the earliest known clear fossils of simple
single-cell life appeared 0.9 billion years after the earth cooled
(4.5 billion years ago), though other evidence suggests life after
only 0.5 billion years [Balter 96]). The earlist known large complex
single-cell fossils ("eukaryotic" in appearance) then appear about
2.0 billion years after this early evidence. 0.8 billion years later
the tempo of evolution picked up dramatically, perhaps with the
invention of sex [Schopf 95], and then 0.5 billion years later we see
the first substantial fossils of multi-cellular life [Knoll 95].
Finally, 0.6 billion more years brings us to where we are today.

While these periods are not exactly equal, they are roughly
consistent with the (roughly exponential) distribution of actual
durations between hard steps predicted by the above model of trial
and error steps. Some important complications and caveats, however,
must be considered.

First, assuming the first step happened on Earth, all we really know
is that it must have happened sometime between when the Earth cooled
enough to support life, and the age of the the earliest known
fossils, which also happen to be the earliest known rocks where one
could possibly see such fossils. Thus all we can say is that this
first step took between 0.0 and about 0.5 billion years. And since
the environment of early Earth was unusual, there may have been a
special window of opportunity within which several discrete steps
took place.

Second, the appearance of the earliest known large complex
single-cell fossils corresponds closely with Earth's transition to an
oxygen-dominated atmosphere, a transition which seems to have been
awaiting the slow oxidation of all the ocean's iron. Since eukaryotes
need oxygen to breathe, they likely could not have been widespread
before this point. Thus a hard trial-and-error step likely did not
happen at this point in time. One or more hard steps might have taken
place before this, however, within populations too small to show up
in the fossil record. The "potential" created by these hard steps
might have required an environmental change in order to "flower".

Third, the famous Cambrian explosion of about 0.6 billion years ago
was also simultaneous with some independent environmental changes,
such as the breaking up of a supercontinent and the end of the
Earth's worst ice age ever. If we think of environmental event as
random, we can model this as a double biological/environmental hard
step: Some biological hard step first created a potential, a
potential which could not be realized without a compatible later
environmental hard step.

Finally, brain size relative to body size has been increasing
somewhat steadily for both mammals and birds ever since the mass
extinction of 65 million years ago (most likely also caused by an
external event such as an asteroid) eliminated the dinosaur
competition [Russell 83, Jerison 91]. Thus if large brains were the
most recent hard step then this step would have to be placed at least
about 0.3 billion years ago, where we find the most recent common
ancestor of mammals and birds soon after the invention of the Amniote
egg (which allowed animals to colonize land) [Ostrom 92].
Alternatively, perhaps the most recent hard step was the development
of a language potential in mammals, and not in birds, a potential
which wasn't exploited until brains got large enough. (Mayr seems to
think birds were not up to the task [Mayr 85]).

Putting all this together, a better guess of the hard steps would be
as follows. First one or more hard steps happened within the first
0.5 billion years after Earth cooled. Then zero or more hard steps
happened while waiting for the ocean's iron to oxidize. Next, one or
more hard steps occurred over the next 0.8 billion years, the last of
which (perhaps the invention of sex or perhaps of archaeatic cells)
finally released the potential to affect the fossil record about 1.2
billion years ago.

A double biological/environmental step then occurred over the next
0.5 billion years to create widespread multi-celled life, and then
0.3 billion years later a hard step of the invention of the Amniote
egg occured. Finally, over the last 0.3 billion years, there have
either been no hard steps, just the steady working out of new
possibilities, or there has been a single or double hard step,
something like the invention of a mammal language potential, which
required a random (but perhaps not hard) environmental event 65
million years ago to begin to be released.

A typical expected hard step duration of about 0.3 billion years
seems a simple fit to this data. And with this fit, it is then
natural to estimate one life hard step at the beginning, then zero to
eight steps leading to complexity, two to three steps leading to sex,
a double step to society, a single cradle step, and then perhaps a
final language step. Overall, we might estimate a total of roughly
seven to nine hard steps here.

This model suggests a number of predictions which may help confirm or
disconfirm it. For example, this model predicts that the expected
time remaining until the window of opportunity for life on Earth
closes is about 0.3 billion years [Carter 83]. This model could
therefore be confirmed by astronomical analysis regarding expected
durations until the Earth suffers a runaway greenhouse effect,
runaway glaciation, too high an oxygen content for land life to
persist, a serious instability in the sun, a nearby supernovae, a
massive asteroid impact, or by some other disaster ahead in the sun's
travels through the galaxy [Barrow & Tipler 86, Leslie 96].

This model also implies that as long as some evolutionary step took
sufficiently long, the actual time taken does not indicate how hard
the step was. Thus we'll have to use other cues to find the hardest
steps among the hard ones. Finally, this model strongly suggests that
our ancestors passed through at most one hard trial and error step in
the last hundred million years. This last step might, however, have
required some special conjunction of features, such as large brains
and good hands, to appear in the same animal at once. (These further
predictions of this model have not been published elsewhere, to my
knowledge.)

To these roughly nine biological hard steps we might add two other
discrete (random but not trial and error) type steps: an initial step
of getting the right sort of planet around the right sort of star,
and a final step of humanity either succeeding or destroying itself
soon. Together, these eleven steps could explain the Great Filter if
the (logarithmic) average filter per step was at least a factor of
one hundred. That is, either there might be, on average, a one
percent chance of passing a discrete step, or about a thirty billion
years expected time to complete a trial and error step. Of course the
Great Filter need not be distributed evenly among these steps - just
how much of the filter rests in the last step is the ominous question
that motivates our analysis.

The recent evidence of simple single-cell Mars life [McKay et. al.
96] is relevant for reconsidering the steps prior to single-cell
life. If there really was single-cell life early in Mars' history,
and if we find that it was different enough to imply that it probably
evolved independently from life on Earth, then unless Earth and Mars
shared some special unusual environment, the total step from dead
matter around the right sort of star to simple single-cells must be
pretty easy. Future optimism would then have to be based on other
past steps.

If life evolved on one of these planets and was spread to the other
via a local panspermia, then we don't know much more than we did
before. But if single-cell life started before our solar system, and
spread here via a wider panspermia [Crick 73, Weber & Greenberg],
then that could help. It would allow there to have been many more
trial and error hard steps taking perhaps ten billion years. This
seems especially plausible given the amazing complexity of the
earliest life we see, and that this life has survived virtually
unchanged to this very day.

This wider panspermia scenario also allows steps prior to our
single-cell life to be more improbable for any one region of space,
but at the expense of making the next step that much more probable,
by providing more places for it to start from. Wide panspermia of
complex single-cell life could also be possible, but seems less
likely given that such life seems less robust to extreme
environments, and more tuned to Earth's environment [Crick 81].

Radio signals from extraterrestrial intelligences would of course be
strong information regarding the size of the entire filter up to the
point where such signals are possible. Not only would this
information help pin down our biological expectations, but it would
also seem to be bad news regarding our explosive future. And the
nearer such signals originated, the worse this news would be (though
see the zoo-hypothesis discussion below). Conversely, negative
findings would be good news, and the prospect of this should
encourage such research. Note this is the opposite of the usual
justification offered by SETI researchers, who usually focus on the
valuable information extraterrestrials might send us.

Research into SETI and the evolution of life does much more than
satisfy intellectual curiosity - it offers us uniquely long-term
information about humanity's future.

Reconsidering AstroPhysics

There are also several ways in which we might reconsider our
understanding of physics and astronomy to help explain the Great
Filter.

One possibility is that fast space travel and colonization between
stars and galaxies is much harder than it looks, and effectively
impossible, even for nanotech-based machine intelligence. The
interstellar medium, for example, may be much harsher than we
realize. This would suggest we have good chances of surviving, but
little prospect of leaving our solar system at any substantial speed.
The slower the maximum speed, the smaller is the Great Filter that
needs to be explained.

Another possibility is that the universe is very much smaller than it
looks, perhaps because of some non-trivial topology, so that our past
light cone contains much less than it seems. This would also reduce
the size of the Great Filter needing to be explained.

Perhaps the most optimistic physics alternative is that it is
relatively easy to create local "baby universes" with unlimited mass
and negentroy, and that the process for doing this very consistently
prevents ordinary space colonists from escaping the area, perhaps via
a local supernovae-scale explosion. The amount of the Great Filter
this could explain would depend on just how consistently such
escaping colonists are prevented.

There are also three "save stellar appearances" astrophysics
alternatives which could explain why an apparently dead universe is
really alive, with our system an isolated "zoo" [Ball 73].

First, large-scale engineering such as orbiting solar collectors made
from asteroids, Dyson spheres, and stellar disassembling might be
effectively impossible, explaining why nearby stars look so natural.
Second, structures that best use such resources might happen to
almost always preserve natural spectra and other appearances. Third,
our understanding of astrophysics might just be very wrong, so that
the apparently dead stars and galaxies around us really are very
alive.

Yet another possibility is that advanced life mainly colonizes "dark
matter", mainly leaving fallow the stars and other ordinary matter we
see. This scenario would require a stronger version of the zoo social
hypothesis, which I call a "common zoo", discussed below.

Our understanding of dark matter as simple dead matter is progressing
rapidly, however, and may soon help confirm or deny this possibility.
Recent gravitational lensing observations [Bennett, et. al. 96]
indicate that about half (and perhaps all) of the dark matter in our
galactic halo consists of objects from one solar mass to one tenth
this, and relatively little is in the range below this down to Earth
size objects. The smallest independent object in this range yet seen,
a brown dwarf of 20-50 Jupiter masses, has an understandable
Jupiter-like spectra [Savage, Sahli, & Villard 95].

Rethinking Social Theories

I personally think that most of the Great Filter is most likely to be
explained by the steps I think we understand the least about: the
steps in the biological evolution of life and intelligence. However,
many physical scientists focus on explaining the filter via the area
they seem to think we understand the least: social science.

Astronomers Sagan and Newman, for example, claim that either we will
destroy outselves with nuclear weapons, or learn to "live with other
groups in mutual respect" by losing "our own predispositions to
territoriality and aggression. ... This adaptation must apply ...
with very high precision, to ... every individual within the
civilization", so that we become the "least likely to engage in
aggressive galactic imperialism" [Sagan & Newman].

Similarly, Papagiannis claims that "those that manage to overcome
their innate tendencies toward continuous material growth and replace
them with non-material goals will be the only ones to survive this
crisis," implying a galaxy "populated by stable highly ethical and
spiritual civilizations" [Papagiannis 84]. And Stephenson claims that
"for a truly advanced intelligence the drive for quality rather than
redundant quantity would be paramount" [Stephenson 82].

Now of course if a substantial fraction of civilizations followed
such scenarios, these theories could explain a small part of the
Great Filter. But to explain a substantial part of the Great Filter,
such scenarios would have to follow from situations like ours with a
very high reliability. While this is logically possible, these
authors offer no reasons for expecting such a situation. These
theories thus seem more like wishful thinking than serious attempts
to explain the phenomena using our best understanding of the social
sciences.

To the contrary, while one expects temporarily powerful groups to
have temporarily stronger tendencies toward both colonization and
combat aggressiveness, controlling for this there is no known
correlation between these factors, nor any known theoretical reason
to expect such a correlation. And even if a one-time event did select
for low colonization tendencies, we would expect stronger tendencies
to eventually be selected back if variation was still allowed.

Social scientists have good reasons for expecting competitive
populations to both generically fill new niches, and to shy away from
wars with severe consequences, and social scientists who have
considered the subject have expected substantial interstellar
migration [Finney & Jones 85].

Given the confusion this topic seems to produce, it seems worth
mentioning that one shouldn't put great hopes on the idea that now
that we have control over genetic processes, intelligence can free
itself of "biological imperatives" and choose new purposes. Crabgrass
does not colonize because it has a purpose to fulfill a biological
imperative. Biological organisms have always been free to pursue
whatever purposes they want, and to invent new ones. The point is
that in general the creatures whose purposes lead to the most
reproduction end up dominating the future.

Similarly, human control over genetics will change the way variation
is encoded, and greatly speed up the variation process, but will by
itself not let humans escape the basic evolutionary process of
variation and selection. Avoiding this process would require global
control over reproduction, implying at least a strong world
government regulating child-bearing, local economic growth, and even
the spread of ideas, with a political process undemocratic enough to
avoid variation and selection working through the political process.

The following social hypotheses, though still seemingly unlikely, are
at least minimally plausible and are at least grounded in our
understanding of social science.

The most pessimistic social scenarios are scenarios like massive
nuclear war or ecological disaster. Such devastating war would likely
need to be prior to dispersal across the solar system, unless it
could destroy our sun. And an ecological failure would need to be
prior to an ability to transcend our biological inheritance, such as
via machine intelligence (uploaded or artificial). It seems possible,
though unlikely, that only one in a million worlds at our stage
avoids such a fate. While even this would still leave most of the
Great Filter to be explained in some other way, the prospect of such
a possibility is a strong motivation for studying the Great Filter.

A related scenario would be some sort of unspecified social collapse,
of the sort that lead to the fall of a variety of relatively isolated
ancient civilizations (such as Easter Island), only much more severe,
so that nothing was left to rise from the ashes and try again. When
we better understand these historical events, perhaps we will be in a
better position to dismiss this possibility.

A devastation scenario is implicit in the usual formulation of the
Drake equation. For prior evolutionary steps the equation asks for
the probability that the system will reach the next step, but at our
level of evolution, the equation asks for the expected time until the
civilization disappears, and once gone it is assumed to never return.

Another approach to alternative social theories is to note that if
our descendants are no longer sufficiently internally competitive,
the evolutionary model need no longer apply. For example, if one is
willing to assume a closed universe and that FTL travel out from an
explosion point is possible, one might hypothesize that the first
civilization anywhere to reach an explosion point happened to have a
strong stable central government (like Imperial China) which placed a
very high ideosynchratic value on preserving the natural appearance
of the universe [Freiheit 93, Crawford 95]. By being first and spread
out very fast, these conservationists might enforce their preferences
on all late-comers.

The FTL could be via a "warp" drive, as in [Alcubierre 94].
Constructed wormholes would not be sufficient to expand faster than
lightspeed, because the hole ends must move normally. Pre-existing
"long" wormholes might be sufficient though.

Without FTL travel, a conservationist scenario would require that a
strong vast majority of civilizations somehow obtain a
conservationist preference, and that a conservationist policy not put
a leave-it-be conservationist civilization at a substantial military
disadvantage to pave-it-over competitors. The average size and
density of non-conservationist powers would also need to not conflict
with our apparent lack of observing such differing cosmological
regions.

No special social theory would be required for a "zoo" hypothesis
[Ball 73] which is bundled with one of the astrophysics alternatives
listed above which would imply that aggressively colonized systems
look just like natural ones. It is natural enough to suppose that
some small fraction of places would be left as nature preserves. One
seems to need a special social theory, however, to explain a "common
zoo" hypothesis, that most all matter visible to us has been set
aside as nature preserve.

The common patterns of visible matter across the observable universe
would have to be explained by a remarkably common preferences for the
density and nature of such preserves, and a common lack of preference
for any visible partially-restructured "gardens". There would also
need to be a remarkably widely coordinated effort to punish deviant
powers who might attempt to send radio signals or self-reproducing
probes to such wildlife preserve stars. Consider, for example, that
the energy of a single star might power an intermittent very
narrow-band signal detectable to pre-explosive life like ours across
the entire universe [Gott 82].

I mention this common zoo hypothesis not because I find it especially
plausible, but because it is among the most plausible scenario I can
construct without also invoking astrophysics alternatives like FTL
travel. It thus illustrates the extremes required to construct
self-consistent purely social explanations of the Great Filter.

Conclusion

No alien civilizations have substantially colonized our solar system
or systems nearby. Thus among the billion trillion stars in our past
universe, none has reached the level of technology and growth that we
may soon reach. This one data point implies that a Great Filter
stands between ordinary dead matter and advanced exploding lasting
life. And the big question is: How far along this filter are we?

To support optimism regarding our future, we must find especially
improbable past evolutionary steps. And in fact we can find a number
of plausible candidates for groups of hard trial-and-error biological
steps: life, complexity, sex, society, cradle and language. Presuming
there are about nine hard steps total here, the Great Filter could be
explained if the expected time for each of these steps averaged
(logarithmically) to about thirty billion years, if only one percent
of stars could support such steps, and if we have only about a one
percent chance of not destroying ourselves soon (or permanently
banning colonization).

While one might also explain parts of the Great Filter via unusual
approaches to astrophysics or social science, such assumptions seem
less plausible to me than thirty billion year expected times for the
identified biological steps. There is ample room for disagreement,
however.

The larger the remaining filter we face, the more carefully humanity
should try to avoid negative scenarios. To inform such choices, we
would do well to analyze all these issues more carefully, and to
collect more relevant data.

Fortunately, rapid progress is being made in several relevant
empirical areas. Dark matter astronomy may soon confirm or deny the
common zoo hypothesis. Mars life evidence may soon indicate the ease
of the earliest steps in evolving life.

Other progress also continues, at a slower but still encouraging
pace. A wide variety of research continues to illuminate the early
history of life on earth. Theoretical physics is closing in on
whether FTL travel is possible. And speculative engineering is
helping to estimate the feasibility of interstellar travel and large
scale solar system constructions. Astronomers and global modelers are
working to evaluate how long the Earth should remain hospitable to
life (if we don't destroy it). And social scientists continue to
improve our understanding of what might effect colonization and
self-destruction tendencies.

It may not be too long before spacecraft can test theories of wider
panspermia, perhaps by looking for single-cell life within comets.
And SETI researchers continue to test the hypothesis that life at our
stage is dense, so that we still face an enormous filter. (They might
also consider looking for common-zoo renegade broadcasters from
across the universe.)

Finally, we would do well to keep a in mind a few unusual aspects of
this Great Filter puzzle. First, let us keep in mind the
interdisciplinary nature of the this puzzle. While it may comforting
for each discipline to claim that the Filter must surely lie in some
other discipline of (in their eyes) lessor repute, such claims should
surely be backed up by detailed analysis using our best understanding
of that discipline. It will no more do for astronomers to simply
claim, without further supporting analysis, that people will lose
their tendency to colonize, than it would do for biologists to simply
declare that astronomers could not possibly know that the universe is
as big as they claim.

Second, we must be wary of the "God of the Gaps" phenomena, where
miracles are attributed to whatever we don't understand. Contrary to
the famous drunk looking for his keys under the lamppost, here we are
tempted to conclude that the keys must lie in whatever dark corners
we have not searched, rather than face the unpleasant conclusion that
the keys may be forever lost.

Finally, we should remember that the Great Filter is so very large
that it is not enough to just find some improbable steps; they must
be improbable enough. Even if life only evolves once per galaxy, that
still leaves the problem of explaining the rest of the filter: why we
haven't seen an explosion arriving here from any other galaxies in
our past universe? And if we can't find the Great Filter in our past,
we'll have to fear it in our future.

---------------------------------------------------------------------

Acknowledgements

I thank the following for their comments: Curt Adams, Niklas Bostrom,
Brandon Carter, John K. Clark, Mark Crosby, Bradley Felton, Eric Watt
Forste, Sean, Hastings, Eugene Leitl, John Leslie, Sean Morgan, Pat
Powers, Hara Ra, Anders Sandberg, Damien R. Sullivan, and Michael
Wiik.

Technical Appendix

This appendix contains a more precise description of our one data
point, and a derivation of the novel results regarding hard trial and
error steps.

Regarding the data point, consider the cumulative probability F(t,dv)
that a given (cosmologically co-moving) volume of space dv will have
contained the earliest origin of an evolutionary path that results in
an explosion (arriving) there by time t since the big bang, and
moving out to colonize and visibly alter most of the visible
universe. (More precisely, let this probability be contingent on the
universe surviving this long in its familiar physical state, rather
than for example suffering a destructive transition to a lower vacuum
ground state.) If these probabilities are independent for small
volumes, the expected number of other explosions reaching here by T =
the age of universe minus one million years ago is at least the
integral of F(t,dv) across the surface of a past light cone starting
from a million year old event in our history. Using a homogeneous
space approximation (surely valid on cosmological scales), so that F
(t,dv) = F(t)*dv, this is:

    Integral from t = 0 to T of 4 pi F(T-t) t^2 dt

Our one data point gives strong probabilistic evidence that this
integral is not much more than one. This implies that F(t) is very
small! For example, if F is time-independent, so that F(t)=1-exp
(-f*t) or approximately f*t for f*t small, then f*T*
(ave-volume-per-star) is not much more than 1/(number of stars in the
visible universe) or about 10^-22.

Now consider N hard trial and error steps which must be completed in
a certain order within a time window W. If the probability that step
i takes less than time t[i] is 1-exp(-f[i]*t[i]) or about f[i]*t[i]
for f[i]*t[i] small, then assuming all f[i]*W are small, the joint
probability density over the various hard step durations t[i] is
about Product[i] f[i], independent of all t[i]. Conditional on Sum[i]
t[i] < W and all t[i] > 0, this distribution therefore treats all i
the same, regardless of f[i]. Thus conditional on success, all hard
steps have roughly the same distribution over durations, regardless
of how hard they are. (For a more rigorous mathematical treatment,
see [Hanson 96].)

References

M. Alcubierre (1994) "The warp drive: hyper-fast travel within
general relativity", Class. Quantum Grav. 11:L73.

John A. Ball (1973) "The Zoo Hypothesis", Icarus, 19:347.

Michael Balter (1996) "Looking for Clues to the Mystery of Life on
Earth" Science August 16, 273:870-872.

John D. Barrow, Frank J. Tipler (1986) The Anthropic Cosmological
Principle Oxford University Press, NY.

Gregory Benford (1981) "Extraterrestrial Intelligence?" Q. Jl. R.
astr. Soc. 22:217.

David Bennett, Kim Driest, Christopher Stubbs, Alex Rodgers, Kem
Cook, Will Sutherland (1996) "Researchers Determine Machos May
Comprise Fifty Percent of Galactic Dark Matter", Press Release, AAS
Meeting, San Antonio, Texas, January 16.

Glen David Brin (1983) "The `Great Silence': The Controversy
Concerning Extraterrestrial Intelligent Life, Q. Jl. R. astr. Soc.
24:283-309.

Brandon Carter (1983) "The Anthropic principle and its implications
for biological evolution" Phil. Trans. R. Soc. Lond. A 310:347-363.

Brandon Carter (1993) "The Anthropic Selection Principle and the
Ultra-Darwinian Synthesis", in The Anthropic Principle, ed. F
Bertola, U. Curi, Cambridge Univ. Press, 33-63.

I. A. Crawford (1995) "Some Thoughts on the Implications of
Faster-Than Light Interstellar Space Travel" Q. Jl. R. astr. Soc.
36:205-218.

Francis Crick, L.E. Orgel (1973) "Directed Panspermia" Icarus 19:341.

Francis Crick (1981) Life Itself: its origin and nature, Simon and
Schuster, NY.

David R. Criswell (1985), "Solar System Industrialization:
Implications for Interstellar Migrations", in Interstellar Migration
and the Human Experience, ed. Finney & Jones, 50-87.

K. Eric Drexler (1992a) Nanosystems, John Wiley & Sons, Inc., NY.

K. Eric Drexler (1992b) "Molecular Manufacturing for Space Systems:
An Overview", Journal of The British Interplanetary Society, Vol.
45:401-405.

Freeman Dyson (1966) "The Search for Extraterrestrial Technology" in
Perspectives in Modern Physics, ed. R.E. Marshak, Wiley, NY, 641-655.

Freeman Dyson (1979) "Time without end: physics and biology in an
open universe" Reviews of Modern Physics, 51:3, July.

Martyn J. Fogg (1987) "Temporal Aspects of the Interaction among the
First Galactic Civilizations: The `Interdict Hypothesis'" Icarus
69:370-384.

Robert Forward (1985) "Starwisp: An Ultralight Interstellar Probe",
AIAA Journal of Spacecraft and Rockets 22:345-350.

Robert Forward (1986) "Feasibility of Interstellar Travel: A Review"
Journal of the British Interplanetary Society 39:379-394. Ben R.
Finney, Eric M. Jones (1985) Interstellar Migration and the Human
Experience, Univ. of Calif. Press.

Ben R. Finney, Eric M. Jones (1985) "Fermi's Question" in
Interstellar Migration and the Human Experience ed. Finney & Jones,
298-300.

Walter M. Fitch, Francisco J. Ayala (1995) Tempo and Mode in
Evolution, Genetics and Paleontology 50 Years After Simpson, National
Academy Press, Washington D.C.

F. E. Freiheit (1993) "The Possibilities of FTL: Or Fermi's Paradox
Reconsidered, 1993, http://www-personal.engin.umich.edu/~fritx/
Ftlessay/essay.html

Donald Goldsmith (1980) The Quest for Extraterrestrial Life, A Book
of Readings, University Science Books.

J. Richard Gott (1982) "Cosmology and Life in the Universe", in
Extraterrestrials, Where Are They?, ed. M. Hart & B. Zuckerman,
122-134.

J. Richard Gott (1993) "Implications of the Copernican principle for
our future prospects" Nature May 27, 363:315-319. Robin D. Hanson
(1994) If Uploads Come First: The Crack of a Future Dawn. Extropy 6
(2):10-15.

Robin D. Hanson (1996) , Working Paper, September.

Ingemar Hansson, Charles Stuart (1990) "Malthusian Selection of
Preferences", American Economic Review, June 80(3):529-544.

Michael H. Hart (1975) "An Explanation for the Absence of
Extraterrestrials on Earth", Q. Jl. R. astr. Soc. 16:128.

Michael H. Hart, Ben Zuckerman (1982) Extraterrestrials, Where Are
They?, Pergamon Press, New York.

Sabastian von Hoerner (1978) "Where is Everybody?"
Naturwissenschaften, 65:553.

Harry J. Jerison (1991) Brain Size and the Evolution of Mind,
American Museum of Natural History, New York.

Andrew Knoll (1995) "Proteroic and Early Cambrian Protists: Evidence
for Accelerating Evolutionary Tempo" in Tempo and Mode in Evolution,
Genetics and Paleontology 50 Years After Simpson, ed. Fitch & Ayala,
63-83.

Thomas Kuiper, Glen David Brin (1989) Extraterrestrial Civilization,
Am. Assoc. Physics Teachers, College Park, MD.

John Leslie (1996) The End of the World: The Science and Ethics of
Human Extinction, Routledge, NY.

Ernst Mayr (1985) "The probability of extraterrestrial intelligent
life", in Extraterrestrials, Science and alien intelligence, ed.
Regis, 23-30.

Ernst Mayr (1995) "Can SETI Succeed? Not Likely", Bioastronomy News,
7:3.

David S. McKay, Everett K Gibson Jr., Kathie L. Thomas-Keprta,
Hojatollah Vali, Christopher S. Romanek, Simon J. Clemett, Xavier
D.F. Chillier, Claude R. Maechling, Richard N. Xare (1996) "Search
for Past Life on Mars: Possible Relic Biogenic Activity in Martian
Meteorite ALH84001", Science, August 16, 273:?.

Christopher Miller (1995) "Cosmic Hide and Seek: the Search for the
Missing Mass" http://w3.gti.net/cmmiller/drkmttr.html

John Ostrom (1992) "A History of Vertebrate Successes", in Major
Events in the History of Life, ed. J.W. Schopf, 119-139.

Michael D. Papagiannis (1978) "Are We Alone, or Could They be in the
Asteroid Belt?" Q. Jl. R. astr. Soc. 19:277. Michael D. Papagiannis
(1984) "Natural Selection of Stellar Civilizations by the Limits of
Growth", Q. Jl. R. astr. Soc. 25:309-318.

Edward Regis, Jr. (1985) Extraterrestrials, Science and alien
intelligence, Cambridge University Press, NY.

Mark Ridley (1993) Evolution, Blackwell Scientific Publications,
Boston.

Dale A. Russell (1983) "Exponential Evolution: Implications for
Intelligent Extraterrestrial Life" Adv. Space Res. 3(9):95-103.

Carl Sagan, William I. Newman (1983) "The Solipsist Approach to
Extraterrestrial Intelligence", Q. Jl. R. astr. Soc. 24:113-121.

Don Savage, Jim Sahli, Ray Villard (1995) "Astronomers Announce First
Clear Evidence of a Brown Dwarf", Space Telescope Press Release No.
STScI-PR95-48, November 29.

Louis K. Scheffer (1994) "Machine Intelligence, the Cost of
Interstellar Travel and Fermi's Paradox" Q. Jl. R. astr. Soc.
35:157-175.

J. William Schopf (1992) "The Oldest Fossils and What They Mean", in
Major Events in the History of Life, ed. J.W. Schopf, 29-63.

J. William Schopf (1992) Major Events in the History of Life, Jones
and Bartlett, Boston.

J. William Schopf (1995) "Disparate Rates, Differing Fates: Tempo and
Mode of Evolution Changed from the Precambrian to the Phanerozoic",
in Tempo and Mode in Evolution, Genetics and Paleontology 50 Years
After Simpson, ed. Fitch & Ayala, 41-61.

Robert Shapiro, Gerald Feinberg (1982) "Possible Forms of Life in
Environments Very Different from the Earth", in Extraterrestrials,
Where Are They?, ed. M. Hart & B. Zuckerman, 113-121.

George G. Simpson (1964) "The nonprevalence of humanoids" Science
143:769.

Peter Skelton (1993) Evolution: A Biological and Palaeontological
Approach, Addison-Wesley, NY. David G. Stephenson (1979)
"Extraterrestrial Cultures within the Solar System?", Q. Jl. R. astr.
Soc. 20:422.

David G. Stephenson (1982) "Models of Interstellar Exploration" Q.
Jl. R. astr. Soc. 23:236-251.

Frank J. Tipler (1980) "Extraterrestrial Intelligent Beings do not
Exist", Q. Jl. R. astr. Soc. 21:267-281.

Frank J. Tipler (1981) "Additional Remarks on Extraterrestrial
Intelligence", Q. Jl. R. astr. Soc. 22:279-292.

P. Weber, J.M. Greenberg (1985) "Can Spores Survive in Interstellar
Space?" Nature 316:403-407.

Paul S. Wesson (1990) "Cosmology, Extraterrestrial Intelligence, and
a Resolution of the Fermi-Hart Paradox", Q. Jl. R. astr. Soc.
31:161-170.

D.P. Whitmire, D.P. Wright (1980) "Nuclear Waste Spectrum as Evidence
of Technological Extraterrestrial Civilizations" Icarus 42:149-156.

O.B. Zaslavskii (1996) "Generalized second law and the Bekenstein
entropy bound in Gedankenexperiments with black holes" Class. Quantum
Grav. 13:L7-L11.

B. Zuckerman (1985) "Stellar Evolution: Motivation for Mass
Interstellar Migrations" Q. Jl. R. astr. Soc. 26:56-59.

---------------------------------------------------------------------

Robin Hanson hanson@hss.caltech.edu September 15, 1996


[Pages Linking Here] known by AltaVista
This page is translated to Belorussian here, and to Irish here here.