Hacker News new | past | comments | ask | show | jobs | submit login
Galaxy Simulations Offer a New Solution to the Fermi Paradox (quantamagazine.org)
169 points by Errorcod3 14 days ago | hide | past | web | favorite | 177 comments



I have a boring solution to the Fermi paradox. It is old, but really, really boring.

Start looking at things that are weird about our Solar System and the Earth. Figure out which ones are probably necessary for intelligence. Add those into the Fermi calculation. And see that it is likely that we're the first intelligent life in this galaxy.

What are some of those special things? It is not enough to have a planet that has the right ingredients to start life. It is necessary that it's progress not be constantly wiped out by major disasters over a very long time scale. What kinds of disasters?

You don't want a supernova going off too close. We have been lucky to avoid that. Our odds were greatly improved by the fact that we spend most of our time out of the galactic plane, away from other stars that could be about to go boom. This is an unusual orbit.

You don't want to be hit by too many comets. We've been hit by some, but far, far less than we would have without Jupiter acting like a sweeper to clean up dangerous stuff in our neighborhood. How rare is that? We have cataloged planets in hundreds of other solar systems. We see ones with only rocky planets. Ones with only gas giants. But we're the only one we know of with both gas giants and rocky planets. The only one where the rocky planets wind up protected from most of dinosaur killer kinds of impacts.

Oh, and get this one. Without the Moon, the tilt of the Earth's spin is unstable. Without it, in simulations we would wind up with one pole aimed at the Sun and the other not once every few tens of millions of years. Probably not good for the development of intelligent life. Again we don't know how rare this is, but we suspect that it is rather uncommon.

Suppose that each of these only happens to one star out of a thousand. Suppose further that there is one more, as yet unidentified, special factor about us that is also required. Again make that a 1/1000 coincidence. That would make the odds good that of the ~250 billion stars in our galaxy, we're the only ones with intelligent life.

Or if you do the back of the envelope just with the three that I named, and made them 1/10,000 coincidences, you get the same result.

As a sanity check, the fact that no other intelligent life has been observed is evidence that it is unlikely. And given Fermi's argument, it is probably unlikely at least on the scale of 100 billion to one against.


We have cataloged planets in hundreds of other solar systems. We see ones with only rocky planets. Ones with only gas giants. But we're the only one we know of with both gas giants and rocky planets. The only one where the rocky planets wind up protected from most of dinosaur killer kinds of impacts.

Right, because that’s what or current techniques mostly allow us to find. Big planets close to their primaries are most favorable for techniques based on transiting such as light curve analysis. Drawing conclusions based on limited data with primitive tools is a poor notion though, and it tells us nothing about the typical makeup of extrasolar systems.


I think you are both forgetting about Venus. And kinda Mars.

I think there are very many things about Earth that are underappreciated as possible limiting factors.

I think that plate tectonics is absolute prerequisite for life to evolve. Not only that, you need to have plate tectonics for a very long time for life to evolve. You need sharp temperature gradients stable for very long time on geologic scales and you only get that with plate tectonics. Without plate tectonics and with presence of atmosphere and voluminous bodies of water (other requirements) everything just gets obliterated quickly and the entire planet becomes flat without much features or temperature gradients. At later stages of evolution I think geographic features are important as they create niches which promote specialization and differentiation.

I am currently reading on origins of political order (https://www.audible.co.uk/pd/The-Origins-of-Political-Order-...). I think it might be possible large brain and high intelligence are not inevitable evolutionary results. The author speculates, and very convincingly, that the intelligence is result of human and chimpanzee males and females having different goals that cause them to deceive each other. The one who could better deceive would have evolutionary advantage and this caused humans to grow in intelligence as a result. But this level of intelligence is not inevitable as we see other species dealing with the problem in different ways. Large brain is a huge liability to an animal as it consumes a lot of energy and causes lots of other issues. It is true that animals use large brains to avoid predators or, being predators, to be more effective at the task. But it also seems that most animals don't seem to be growing their brains more with relation to their body size as it seems there are diminishing returns at some point and bigger brain costs exponentially more than it is worth.


Nothing in evolution is inevitable. Granted some things are more likely (oxygen-binding hemoglobin has independently evolved several times from its related non-binding globin family). Brains, nervous systems, even multicellularity are not inevitable. The specific combination of astronomical, geological, and biological features that lead to our development might not lead to similar outcomes if ran a million times.

As other responders have pointed out, brains are just one possible substrate for intelligence. We can't account for other systems (biological or otherwise) that could lead to the development of intelligence, so of necessity any probability estimate based on us has to be a lower bound of possibility.

One bit of our earth history that is worth noting is that the conditions when life presumably arose are hostile to multicellular life as we know it. Earth's early oceans were hot, anoxic and reducing, and under heavy pressure from far thicker atmospheres. The great oxidation event essentially turned earth's atmosphere into a bacterial waste dump, the waste being oxygen produced by early cyanobacteria. Life had to evolve ways to cope with oxygen or learn to hide. This is still true of our own cells. While we need oxygen to survive, it is still and will always be toxic. Free radical damage such as from reactive oxygen species is a constant threat, and is the ultimate form of irrecoverable information death when we die.

Might it be that the conditions under which life can arise, and under which intelligent life can arise, are different? We got lucky that earth traversed from the first to the second. Add this to the list of a million things that had to go right for us to exist at all. The same might well be true of any other intelligences out there. And the odds of two arising close enough in time and space to contact each other would be far less common than lonely civilizations looking out into the dark.


> The great oxidation event

It took 2 billion years for the hard-working unicellular cyanobacteria to produce enough oxygen to first oxidize the rocks and minerals on the Earth's surface, and to build up enough oxygen in the atmosphere for the birth of multicellular life to be possible.

But if the planed had only supported one half of the cyanobacterial biomass it did, or if the amount of rocks to oxidize, or the mass of the atmosphere, was larger, that 2 billion years could have as well taken e.g. 4 billion years. And the Sun would have already become too hot, and multicellular life would never have evolved.

https://en.wikipedia.org/wiki/Great_Oxygenation_Event


I think there is no way we can know what is necessary for intelligence.

Life is not necessary for intelligence. As a counter-example, intelligent robots are certainly possible. What other substrates might support intelligence?

Life does not require proper axial tilt or distance from supernovas. There is life on the bottom of the ocean that survive on the heat and chemicals from hydrothermal vents, a metabolic pathway that does not depend on the sun. What other possibilities are there that don't happen to be present on earth?

We are biased towards human size, time scales, and biochemistry. And time and again, we have been surprised by forms of life that work differently.


Optical computing fascinates me. Imagine an intelligence based on that.


There was an argument somewhere that biological computational substrate density (with respect to round trip signal speed) was critical.

If you've only had time for 3 thoughts (to simplify) since evolving intelligence... a civilization of your beings probably isn't doing too well.


The most likely solution to the Fermi Paradox is probably the most depressing for the people who are interested in it: space colonization is basically impossible.

The fastest object we've sent outside of the gravitational field of the Sun would take about 75,000 years to reach the nearest star (if we pointed it at it). For perspective, the last glacial advance started in earnest about 75,000 years ago. We don't have any technology that can maintain an automated craft for that long, let alone life support for sentient creatures. Hell, our current technology is insufficient to permit a self-sustaining colony on any other planet or satellite.

Ultimately, the Fermi Paradox is based on the assumption that, given enough chances, an unlikely event is basically certain. While this is true, to do a proper analysis, you do need to look at orders of magnitude to see if there really have been enough chances. And we have so little information to base the computations on that the dominating term is your priors: if you think we ought to have seen alien life by now, you'll pick priors that suggest life is quite likely and find that we ought to have seen it; if you think we oughtn't, your prior will reflect that.


It's worth looking at Project Orion:

"Later studies indicate that the top cruise velocity that can theoretically be achieved are a few percent of the speed of light (0.08–0.1c).[17] An atomic (fission) Orion can achieve perhaps 9%–11% of the speed of light. A nuclear pulse drive starship powered by fusion-antimatter catalyzed nuclear pulse propulsion units would be similarly in the 10% range and pure Matter-antimatter annihilation rockets would be theoretically capable of obtaining a velocity between 50% to 80% of the speed of light. In each case saving fuel for slowing down halves the maximum speed. The concept of using a magnetic sail to decelerate the spacecraft as it approaches its destination has been discussed as an alternative to using propellant; this would allow the ship to travel near the maximum theoretical velocity.[18]

At 0.1c, Orion thermonuclear starships would require a flight time of at least 44 years to reach Alpha Centauri, not counting time needed to reach that speed (about 36 days at constant acceleration of 1g or 9.8 m/s2). At 0.1c, an Orion starship would require 100 years to travel 10 light years. The astronomer Carl Sagan suggested that this would be an excellent use for current stockpiles of nuclear weapons"

https://en.wikipedia.org/wiki/Project_Orion_(nuclear_propuls...


It obviously dependents on priors, and figuring out realistic ones is kind of the point. We haven't observed other life in the universe, so any result of adding up all those chances necessarily has to be low. The question is what makes it so low. And most importantly: Is it a check we humans already cleared or something still coming? That should be a huge warning to us.

> We have cataloged planets in hundreds of other solar systems. We see ones with only rocky planets. Ones with only gas giants. But we're the only one we know of with both gas giants and rocky planets. The only one where the rocky planets wind up protected from most of dinosaur killer kinds of impacts.

Not true: https://en.wikipedia.org/wiki/Kepler-90


Thank you for that correction.

I am having trouble finding the reference that I found that claim in. But note that Kepler-90 does not have rocky planets that look like the Earth. It has super-earths, and gas giants. A gas giant will do a worse job of protecting a super-earth than it will of protecting small rocky planets.

So Jupiter still seems rare.

Incidentally while looking for the article that I saw before, I stumbled across https://www.reddit.com/r/askscience/comments/4ohkkn/how_uniq... which pointed out something I had not thought of. Our Sun is an unusual combination of old and metallic. It is likely that a neutron star collision happened near the stellar nursery where the Sun is born. This is important for a couple of reasons. The first is that metallic stars give rise to more planets. The second is that metallic stars can give rise to planets with richer chemistry. This is another way in which we are weird, which could be potentially important for the development of intelligent life.


Statistics suggest that there is at least 1 planet per star. Multi-planetary systems are the norm, not a rarity.

There is a huge amount of observational bias in exoplanet research and given that we've "only" found a few thousand planets, it's naive to think our Solar System is unique. It's practically certain that with better instruments we'll find new planets in currently known systems.

For example HD_10180 (7 planets currently) also has two potential gas giants. See also HD 219134.

Also: stats of 100 billion to one, those are reasonable odds. That would make two species in the Milky Way, let alone the other billions-trillions of galaxies in the Universe.


I suspect that you almost exactly reversed the point.

The point of the Fermi paradox is that an advanced technological system should be able to colonize the whole galaxy in a geologically short time scale using Von Neumann machines. So if it had happened anywhere else in the galaxy, we'd expect to know about it because they would be already here. But they are not here. So we're probably the first in this galaxy.

Given that there are around 250 billion stars in the galaxy, if the odds were above 1/100 billion per star, the observed outcome would be unlikely. So it is probable that the real odds are below 1/100 billion.

I was offering potential requirements under which real odds on the order of 1/trillion could be justified. These odds are low enough that the observation "we're the first technological civilization in the Milky Way" is not a surprise and the Fermi paradox stops being a paradox.

Neither the Fermi paradox nor my argument suggests that there aren't lots of other galaxies which contain advanced technological civilizations. The time for an advanced galactic civilization to cross intergalactic space is NOT a geological eye-blink. And almost all galaxies that we see are in other clusters, which are much, much farther off.


I think the Von Neumann machine bit is a vastly larger issue than people assume.

The Milky Way is 105,700 light years across. At voyager probe speeds that’s ~1.8 billion years. Stopping requires 4x as much energy, and cutting that time in half requires 4x as much energy. Worse slowing down to build a probe adds more time as does however long it takes to select targets, gather resources, and build several more probes.

On top of a huge number of issues could slow things down such as lost probes, or needing to avoid the center of the galaxy resulting in an even longer trip. Not to mention collisions, at a minimum you need to survive grain of sand sized collisions, but interstellar space may contain significant numbers of much larger objects.

Finally these probes would both come here, and stick around for us to see. A true space based civilization might have little interest in planets like ours, or have sent multiple probes to our solar system that failed.

It seems possible for hundreds of civilizations to all be expanding across the galaxy even if we have seen none of them.


Using https://en.wikipedia.org/wiki/Nuclear_pulse_propulsion we could, with currently conceivable technology, build spaceships that travel at several percent of the speed of light, and decelerate when they reach their goal. At 1% of the speed of light, we can cross go 100,000 light years in 10 million years. This is sufficiently short relative to the time that it took intelligent life to evolve that it would be a pretty good coincidence if there were two that just happened to reach that stage that close together.

Note that with the magic of exponential growth, a few probes sent long distances and then spreading out locally should be able to reach every star in a galaxy with surprisingly little extra time over simply making the trip.


Nuclear pulse propulsion might be possible and it might approach the initial assumptions, but unless your talking ounces of cargo reaching and slowing down from 1% light speed is not feasible. We can’t rely on hand waving science fiction, they would need to send enough infrastructure to booststap manufacturing of a replacement probe. Let’s lowball that at 1 Billion tons of cargo and run through the rocket equation for 10,000 ISP.

Even hand waving that we don’t have force fields to survive single gram impacts at 1% light speed. At those velocities anything is going to act like a massive shaped charge. Space is fairly empty, but when you start talking about the volume swept up across light years that’s really not a safe assumption.

70 years ago people generally assumed fusion was right around the corner. It’s only when you start building these things you get some idea of all the real limitations.


Sanity check. https://en.wikipedia.org/wiki/Project_Longshot was a NASA design that could use nuclear pulse propulsion to deliver 30 tons of payload from here to Alpha Centauri at an average speed of 4.5% light speed. That is with acceleration and deceleration. The same technology could let us go much farther - just with coasting in the middle.

Now it is absolutely true that we could not survive any significant impacts at that speed. However we also have data indicating that the interstellar medium is essentially empty.

Here is a sanity check. Per http://cdms.berkeley.edu/Education/DMpages/FAQ/question36.ht... the density of dark matter (that is, all the stuff that we can't see) is around 6x10^(-28) kg/cm^3. Since you arbitrarily said a gram, let's switch that to g/m^3, and we get 6x10^(-19) g/m^3. The distance to Alpha Centauri is on the order of 4x10^16 m. So a 1000 square meter surface should run into about a gram of "stuff". According to our best theories, most of it is currently unknown stuff, 99% of the rest is single atoms, and the vast majority of the rest is dust particles with sizes on the scale of waves of blue light.

Dealing with that dust and hydrogen is a serious issue. But the odds of running into pebble sized objects at random are pretty low. Heck, even microbe sized objects will be vanishingly rare. True, it will be a disaster if we do. But that translates in practice into, "There will be a failure rate for probes."

The size of a probe to bootstrap manufacturing seems likely to be a lot smaller than a billion tons of cargo. Don't forget, this is going to be encoded using technology that is as far beyond us as we are beyond building the pyramids. I think we can assume rather efficient storage/bootstrapping mechanisms to be available for us.


Work out the kenetic energy of a 0.0001 gram strike at 4.5% c. Ignoring relativity it’s 9 x 10 ^6 J. By comparison, one kg of tnt is 4.184×10^6 J so that’s over 2.1 kg of TNT. Now reolize it’s much worse than that much TNT as it’s penetrating into your shield before releasing that energy. Even atom sized impacts are an issue at those speeds https://en.m.wikipedia.org/wiki/Sputtering

As to your sanity check, Alpha Centauri our closest star, if we are sending multiple probes then you need to targets much further out. Especially if we are trying to get somewhere approaching a straight line in 3D space. The 20th closest star system is 12 light years from earth and on the order of distances we need. Though this gets worse if you need specific kinds of star systems.

That’s pushing 3+ grams under your estimate and your 30 ton payload would need truly massive shielding. Or, you can have a 30 ton shield being activly maintained in front of a postage stamp sized cargo and possibly survive the trip.

As to that billion tons of cargo, it’s almost unimaginably better than current tech could do. We are talking either self sustaining habitats for hundreds of people or safe replicating near full AI on million year timescales. It’s got to gather energy and resources, build CPU’s, engines, etc to gather even more resources. It’s got to travel hundreds of years meaning redundancy and self repair on all critical systems etc etc the list just keeps going.

In terms of engendering from ancient Egypt to now the tallest structure is just under 6x as tall. Some things have improved vastly more and others even less. We simply can’t assume magic even on 10,000 year time scales.


PS: Longshort also used intercial confinement nuclear Fusion which we are no where near getting to work even in ideal laboratory conditions. It’s got a host of it’s own issues and is well into sci-fi territory, as in it might be physically possible but it seems unlikely.

Well your statement was:

> As a sanity check, the fact that no other intelligent life has been observed is evidence that it is unlikely.

Plus some claims about the uniqueness of our solar system which are not necessarily true, given the small sampling of planet's we've found.

There is so much bias in what we observe that I don't think you can make those statements, whether or not you're implying that a species would be galaxy-conquering.

Given these biases I think the odds may well be much better than 100 billion - though I'm not current enough to judge whether that's an appropriate estimate based on curent astrophysics. There's so much uncertainty that it could be orders of magnitude different in either direction.


Jupiter is hard to find. It’s orbit is ~11 years making it harder to measure and more time consuming. It’s pull on the sun is a lot harder to measure than a hot Jupiter, making the radial velocity method more difficult. We aren’t finding average systems. We are finding systems that are easy (and quick) to find using the methods we have available.


Harder but not beyond us. Per https://physics.stackexchange.com/questions/230973/how-can-w... at this point we would be able to pick up Jupiter and Saturn around another star.


Yes, but I don’t see how that is relevant. Your claim is that we aren’t seeing Jupiter/Earth type systems and so they are rare. The counterargument is that we aren’t looking for them, so we don’t see them. We don’t know whether they are rare or not.


Keep in mind that there's many more orbits that don't occlude a star than those that do. So there's some luck required, and likely at least 2 orbital occlusions (22 years) to be sure.


Our understanding of exoplanets is analogous to a 3 month old toddler exploring the world. It's _way_ too early to be making definitive statements about it.



There are so many impossibly complex factors and unknowns regarding how life and the universe works that this type of speculation seems odd to me. The reality is, we don't know how life would or could adapt to any of these differences, or how likely other differences or anomalies we are simply unaware of might have countered or exacerbated various knock-on effects. It's simply guessing.


It is. But it's a very important guess.

In the same way game theory makes guesses about nuclear standoff strategy.


> As a sanity check, the fact that no other intelligent life has been observed is evidence that it is unlikely.

I'm not sure this is particularly useful. It could be like a medieval European saying land to the west of the Atlantic doesn't exist, because they've never once seen it through their looking glasses.

How far could we detect intelligent life, say on a similar level of current humans? If it is based off of radio ways, we would only be able to detect life within a 100 light year radius of earth, which is merely 2% of the way across our galaxy.


Doesn't this boil down to an even more specific version of an 'anthropic principle' argument with a good deal of sampling bias added and easier-to-spot problems? For instance, there's no reason to assume the only way to get a long-term, stable-enough environment for the development of intelligent life is precisely a planetary system with a gas giant (why does it have to be a gas giant?) and a rocky inner planet with a chonker of a satellite.


This is a version of an anthropic principle argument.

However it is not arbitrary, and not an assumption. When you run simulations you can test different combinations of factors and see what difference they make. See, for example, https://www.space.com/31577-earth-life-jupiter-saturn-giant-... to see how important gas giants are for blocking large impacts on small rocky inner planets.

That said, there may be another way to get long-term stability. But if so, then we haven't found it. And yes, we've looked.


and not an assumption.

What do you mean? 'You have to replicate the solar system to come up with intelligent life' is a gigantic assumption, on top of the 'solar system-like planetary system are rare' assumption. Is this a conceivable narrative for the Fermi paradox? Sure. But it's hardly supported enough to make a reasonable probabilistic argument out of it.


What I mean is this. The statements "a large moon makes the axial tilt stable" and "a gas giant greatly reduces how many comets hit small rocky worlds in a solar system" are both verifiable. They are not assumptions.

We also have reasonably compelling evidence that both factors are important for intelligent life to arise. Or, more precisely, good reason to believe that extreme axial tilts and cometary impacts are really, really bad for the biosphere. So the importance of those factors is itself not exactly arbitrary.


You can definitely narrow the Fermi paradox to 'earth-like life' so you can at least attempt to make some sort of estimate. That's reasonable. What's not reasonable is saying this huge assumption is not an assumption. It's totally an assumption, even if it's an assumption you are making because without it, there is even less to say.


There are significantly fewer planets that have gas giants around compared to the ones that don't. It it certain that gas giants prevent debris attacking small rocky planets.

It is obviously an assumption other intelligent life will be like us, we do not know of any other intelligent life to be able to say otherwise. But if I'm looking for life, especially intelligent life, I will look for planets having conditions similar to ours. This is because our sample size is just one, us.

It could be that NOT having gas giants around is how intelligent life ACTUALLY exists in 99.99% of the cases of intelligent life in the milky way galaxy and Earth is the 0.01%. But if I have looked a lot and still just have Earth to guess with, I will continue looking for Earth-like planets. There is no way for me to actually know, but I don't want to waste my time.


There are significantly fewer planets that have gas giants around compared to the ones that don't. It it certain that gas giants prevent debris attacking small rocky planets.

I don't think we know the former and we kind of sort of maybe know the latter for the Solar system right now. As to the rest, I don't follow the relevance, we're not talking about searching for intelligent life at all.


Do you think we're searching for unicellular life then? We're doing our best to search for intelligent life. That means looking for what we know to be able to support life and to eliminate other random searches. Occam's razor and all that. What else are we talking about?

You're right about the former not being certain, but it's definitely rare amongst the planets we've found? And the latter is quite certain.


'planets we've found' is a huge sampling bias because of our current methods. We can't even look for Jupiters, it's unsurprising they are 'rare'.

What else are we talking about?

I mean this thread was not about how we search for life. It's about an estimate of the prevalence of intelligent life - an 'explanation' for the Fermi paradox. I'm saying it's basically a story.


gas giants dont eject material as often and tend to be large gravity wells because thats the defining characteristic that maintains cohesion? (giant)


As often as what? Plus, we've seen some pretty huge rocky planets. Another way of looking this is that it's akin to erroneously calculating the probability that someone might win the lottery twice by multiplying 1/gazillion by 1/gazillion and concluding it can't realistically ever happen.


You don't want a supernova going off too close. We have been lucky to avoid that. Our odds were greatly improved by the fact that we spend most of our time out of the galactic plane, away from other stars that could be about to go boom. This is an unusual orbit.

Not really. The amplitude of the Sun's vertical motion is about 100 parsecs (~ 300 light years), but since the main ("thin") stellar disk has a vertical exponential scale height of about 300 parsecs, the local density of stars at the top/bottom of the Sun's orbit is only about 30% less than when the Sun is crossing through the midplane. (And this isn't at all an unusual orbit.)

What makes a real difference is being relatively far out in the disk from the galactic center: if we were ten times closer to the center, the local density would be almost 100 times higher, and you might indeed have to start worrying about nearby supernovae.


Civilizations probably ended because it's hard to control their own populations. Imagine a society where population is set to 5 billion, the limit only increased when capacity is increased.

This is compounded by:

1) a lot of planets are not hospitable for a variety of factors. A society might resent the limit for many years if it was possible to implement such a limit.

2) the technology era is short-lived and not ongoing. A new civilization starts 10,000 years later. How will they even know homo-sapiens existed if a plague wipes out homo-sapiens tomorrow? A few tiny satellites in the vast emptiness of space?


It'd be ironic if the great filter was that "population required for development and construction of interstellar travel" is very close to "population at which civilizations exhaust their resources and collapse."


The fact that we have never observed other intelligent life does not it doesn’t exist. The unimaginable size of the universe leads me to believe that our special set of conditions are possible elsewhere, though it might be exceedingly rare. And because of this, we will never know about it anyway. You might say that’s the same as it not existing at all, but that’s a question for philosophers braver than I.

One of the scariest thoughts I have ever had is that we truly are alone, but the sheer size of the universe makes me feel slightly better about that prospect.


My idea is we are not alone, never have been and never will be. We are the extraterrestrial big brains that have developed intelligence over millions of years. We long for the others that continued to travel while our ancestors chose to stay and inhabit this planet and in doing so in time forgot about the others and our chosen mission.

> One of the scariest thoughts I have ever had is that we truly are alone

As someone quite new to this discussion, can I ask why? My only immediate reaction is a mild one going the other way: that being alone is reassuring as it means there is not this extra potential threat out there.


On top of all this there's the theory that gas giants typically migrate towards their star, it may be rare for a gas giant like Jupiter to find a stable orbit where it can shield the habitable zone

Another thing making earth rare is its high proportion of iron ore. That and our abnormally large moon are thought to be the result of Earth colliding with another large rocky planet. The tidal force of an extra large moon + our extra ironey core give us an extra strong magnetic field to keep the nasties out and hold our atmosphere



The problem with the drake equation is that we're not really working with means of normal distributions as estimates, we're working with wildly different, skewed distributions for each factor, due to uncertainties spanning orders of magnitude for some factors. Instead of multiplying point estimates you need to convolve the distributions, which yields quite different results.

https://arxiv.org/pdf/1806.02404.pdf


Interesting, thanks for the link


Seems like you got it backwards. Being outside the plane of the galaxy INCREASES the threat from supernova, not decreases it. The further from the plane the less gas/dust/stars etc between you and ever other star in the galaxy.


Um, no.

Intervening material does very little to cut out light. If it did, we couldn't see distant stars.

However the intensity of radiation falls off as the square of the distance from the source of the radiation.

Therefore the primary determinant of risk is the number of stars that could be about to go boom within the critical distance of the Sun. If we're in a less densely populated neighborhood, then our odds of seeing a dangerous supernova go down. Full stop.


Er, there's a giant cloud of dust called the dark rift that blocks a fair bit of the center of the galaxy. We only recently started looking though it at xray wavelengths to see the center of the milky way (and the black hole there). On a clear night you can see the edges of the milky way and the giant cloud of dust.


That cloud is about 2000 light years away from us.

The fact that there are patches of dust doesn't change the fact that we are seldom in the middle of such patches.

And even when we are, it is a lot like the difference between being in light mist versus seeing it from a long ways away. We are, in fact, in the middle of a patch of "mist" called the "local fluff". See https://en.wikipedia.org/wiki/Local_Interstellar_Cloud for more. This does not noticably cut out light between us and stars that are under 100 lightyears away.


Right. This has been my preferred solution to the paradox as well, for the last several years. More and more evidence seems to point towards the Drake Equation simply not sufficiently accounting for uncertainty [1].

There are a few planets and moons in our system that have or had the potential for life, but upon which there's been no evidence for intelligent life, if any life is to be found on them at all: Mars, Europa, Ganymede, among others. While this would seem to increase the probability of life in other solar systems, seeing observable life on only one of the worlds in our system really has a negative impact on the overall odds of life in the universe.

Whether a moon is necessary or not for life on Earth [2], it remains that a stable tilt is currently thought to be critical for the development of complex life. We also had the benefit of an asteroid that obliterated the dominant species of the planet and changed the environment sufficiently to allow for new species to take their place. Hundreds of millions of years of flourishing life also provided us with the abundant, cheap energy sources in the forms of coal and oil that we needed to build civilizations.

There's also an open question as to whether evolution necessarily leads to intelligent species. Earth had large, complex lifeforms for 165 million years, and none of them appear to have developed civilizations. Human societies have existed for just 50,000 years if you take a very liberal definition of "society". Research has found an evolutionary fluke that fed lots of oxygen-rich blood to hominid brains [3]; other research has suggested that intelligence is not only not an evolutionary advantage, it's a disadvantage, due to the much higher calorie diet required to sustain it.

And finally, our universe may not have been hospitable to advanced civilizations until only (geologically) recently. It took a billion years for the universe to form galaxies and stars and planets [4]; it was a more chaotic, hot, radiation-rich universe before our solar system was born. Any civilizations which developed much before our own solar system may have met a premature end by radiation or cosmic debris.

Given the empirical lack of observable life in the universe so far, and that continuing research keeps growing rather than reducing the number of special cases that led to human civilization on Earth, it seems likely to me that even if there are other intelligent civilizations in the cosmos, they are approximately as young as we are, too far away and technologically primitive to make contact.

[1]: A good paper on this from last year: https://arxiv.org/pdf/1806.02404.pdf [pdf]

[2]: https://www.astrobio.net/news-exclusive/the-odds-for-life-on...

[3]: https://theconversation.com/how-our-species-got-smarter-thro...

[4]: https://en.wikipedia.org/wiki/Chronology_of_the_universe


> There's also an open question as to whether evolution necessarily leads to intelligent species.

Especially when, in this context, we're talking about a rather peculiar type of intelligence, ie. the propensity to construct spacecraft and broadcast high-powered message-carrying radio signals. It seems very difficult to argue that there's any evolutionary pressure in favour of these behaviours, just given the time-lag from the development of our large brains to this.


You also need an ocean.


imo you also need dry land, for the tool-making flavor of intelligent life


Naive question: why is that?


It seems unlikely that an aquatic species could ever develop ceramics, metalworking, or any other technology requiring the use of fire. Any ocean covering an entire planet is likely very deep, making construction difficult or impossible. Early power sources such as coal would also be unavailable.


My favorite solution: The galaxy is teeming with life, but in the form of quasi-immortal software that lives at such vast timescales, they're as uninterested in talking to us as we would be in talking to a fruit fly. (Imagine Space Ents that say hello, and are perfectly content to wait 10,000 years for a reply.) It may well be that becoming such a life form is the only practical way to deal with the speed-of-light constraint, either in travel or communication.


We don't have conversations with paramecium.

If you wanted to build a Dyson sphere you would capture all the light from your star and might not exhibit ANY radiation.

We wouldn't see such a species and if there's such a small blink of time between the industrial age to the singularity the probability that there is an overlapping species is very very low.

We would only have about 20-40 years to actually have the technology to find them before we hit the singularity.

We have NO idea what happens after that point but my plan is to just chill with my wife in a virtual Yosemite park once we hit the singularity.


> might not exhibit ANY radiation

That would be against thermodynamics. Every object that has a temperature (above 0K) emits radiation.


You're right obviously, but a Matrioshka Brain would put out such a small amount of radiation, it'd be invisible for all sensor technology outside of, say, black hole lenses.


> a Matrioshka Brain would put out such a small amount of radiation, it'd be invisible for all sensor technology

A huge object emitting infrared radiation would be quite detectable.

At the level of energy output from our Sun, a Dyson sphere with a diameter of the Earth's orbit, would actually have a surface temperature of 395K or 122C. It would put out lots of infrared radiation, and would be very detectable against the cosmic background which has a radiation temperature of 2.7K.

Some SETI programs have looked for these, and already in 1984 we had the technology to observe, and as it turns out, to rule out any of these (Dyson spheres around stars with energy output comparable to our Sun) existing within the nearest 300 light years.

https://www.aleph.se/Nada/dysonFAQ.html#HEAT

https://www.aleph.se/Nada/dysonFAQ.html#OBS

http://www.projectrho.com/public_html/rocket/weirdastronomy....


>We don't have conversations with paramecium.

If paramecium became advanced enough to form a society and want to say something to us, we'd be extremely interested in communicating with it, I think.


How interested are most humans in communicating with the other reasonably advanced species on our own planet?

>once we hit the singularity

I am charmed to read someone here expects this to potentially happen in our lifetime. I know it’s not a completely uncommon sentiment in these circles but I’m not informed about why people hold it - personally, I wouldn’t mind being convinced of it myself. Do you expect us to reach the singularity in our lifetimes, if so why?


Offhand, from the science fiction I've read Dyson spheres make more sense as military installations for the reason you list.

From the perspective of an easier and safer engineering problem even 'wide' "ring worlds" make more sense; those also have the option of using the start inside as a drive for travel. (Probably not very fast, but much like interstellar freight class travel)


> might not exhibit ANY radiation.

They'd still radiate it out, unless they converted it into matter. E=mc² works both ways.


To continue the speculation, such "life" may be what can travel between stars but life such as our may still be common. Indeed, life such as our may be common enough that interstellar life merely takes note of it and moves on since novelty wears off quickly.


Like some British ship visiting some pacific island circa 1800 "Yup, just some birds, palm trees and nothing particularly useful here, note the location in the log and move on."


To me, any solution that requires all alien life acting the same is just as paradoxical as it not existing.


I'm not proposing that such life forms would choose to act the same, but rather that they would be adapted to the constraints of their environment (the speed of c). (There are countless ways to adapt to the conditions of the deep ocean, but none of them are going to be jogging or flight.)

That said, you're right that it's not a given that electronic life would be uninterested in us, even indirectly, such as long-distance observational probes. Strictly speaking, we don't know for sure that that isn't the case. ;)

It's also worth considering that a particular long-lived culture or life form might achieve long-term stable hegemony and/or territoriality, resulting in a consistent policy with regard to either our planet, or planets with organic life in general.

All just speculation, at any rate. I don't claim the model is necessarily correct, just that it's my favorite. :)


I've encountered an idea similar to this:

> The transcension hypothesis proposes that a universal process of evolutionary development guides all sufficiently advanced civilizations into what may be called "inner space," a computationally optimal domain of increasingly dense, productive, miniaturized, and efficient scales of space, time, energy, and matter, and eventually, to a black-hole-like destination.

https://accelerating.org/articles/transcensionhypothesis.htm...


Would you mind dumbing that down for me?

As I understand it we'd be living inside a simulation. Isn't that what GP suggested?


If I grokked the parent right, it's not that we are living in a simulation, but that the apex of development for a species is effectively becoming a simulation.

Now, whether that is an argument that life at present is itself a simulation is yet to be determined. However, I'd imagine given that ungodly level of computational power we'd have the wherewithal to construct a reality for ourselves far more compelling than this one. ;)


We're obviously not living in a best-of-all-possible-worlds simulation, but that doesn't preclude us either from being in a simulation or from being in one created by thoughtful entities for morally good ends.

If you consider that bringing into being all possible worlds and all possible creatures whose existence is net positive in utility to themselves, is a good and desirable and interesting project, then you would end by creating an untold number of simulations, many of which would not be the, or mistakeable for the, best of all possible worlds.

And it would be possible that many creatures who grew up in and formed a unique, not otherwise achievable, identity in the not-perfect worlds could graduate/ascend/sublime into one or more of the more perfect worlds after they had lived a life in their home simulation.

And being such a creature in such a home simulation, not yet ascended or aware that there are other places to ascend to, could also be a solution to the Fermi paradox: we see no other aliens because we have been placed here at least in part to develop our specifically human nature, without the dilution of that nature via the influence of other creatures. Cf. the Star Trek policy of non-intervention, the Prime Directive [https://en.m.wikipedia.org/wiki/Prime_Directive].


Wouldn't it be more fitting to call it a godly amount of computational power then? :)

We're only looking at it from the inside perspective. Who knows, maybe our simulation overlords (ourselves?) already placed us in a far better place than they ever had!


Haha. Well said.

That's a clever take. I suppose if the "original" humans came from a universe as hellish as Warhammer 40K, ours would seem like a cakewalk!


One thing to keep in mind is that ~80% of the universe is 'dark energy' and ~20% is 'dark matter'. The stuff we're made of is ~5% of the mass-density of the universe.

For the stuff you and I are made of, we have a pretty good clue as to what is going on, how it works, what the 'bottlenecks' are in terms of talking with other stuff that is like us.

For dark matter, well... it falls down. That's about all we got. How it interacts, if at all, is a total mystery at this point. And there's about 4x more of it than of the stuff you and I are made up of (which is mostly in stars and black holes, at that!). Does it have phases? Can it have a temperature? If it can sustain something like life, what would that even be like?

For dark energy... um... it falls up? Honestly, we got nothing at this point. We've not really even got an idea if it's a particle or a field or what. Most 'serious' physicists just ignore it as it's embarrassingly mysterious. If life could exist in something like dark energy, we'd be totally left out of any avenue to talk to it. At least with dark matter aliens you can send gravity waves at them. With dark energy ones, we have no idea at all. And dark energy has 16x the mass-density that we do.

My point is that the majority of the universe that we know about today is just a total mystery. And we've only discovered this issue in the last 50 years or so. Meaning that in another 10,000 years, we're likely to have a completely different understanding of the universe than we have today. Like, so radically different that it's akin to the flat-earth model vs. general relativity.

When it comes to alien communication, be patient. It's likely to be so outside our current understanding that it would baffle anyone currently alive; it will be literally unbelievable to you and I.


Interesting, and the probabilities seem to support it. This also means that I'm not going to get to understand any of it in my lifetime, though. _sigh_.


> talking to a fruit fly

I believe it would be very interesting, scientifically, to be able to 'talk' to a fruit fly.


Once. How different can the experiences be between fruit flies after that first contact, though?

And for that matter, what if your species descended from something similar to a fruit fly and you had the processing power to call those species memories to the front of your own mind?


I suspect one of the more interesting aspects of humanity would be that our "day-world" (the sun) is about the same size as our "night-world" (the moon) from our perspective. Sometimes our night world covers our day world completely.

Sun worship might be common, but the human spin on it might be rare.

Studying the impact on our mythology and philosophy could prove interesting to outsiders. It is possible that our species thinks in an extremely binary way compared to other intelligent beings.

Talking to us about it would likely prove less interesting. Betting quatloos on outcomes might prove an interesting sport though.


I don't think you intended your first remark to sound like this, but I'd imagine the same could be said of us from the perspective of a highly advanced interstellar species: how different can the experiences be between humans, right? However, given the level of complexity of a fruit fly I know this metaphor doesn't necessarily hold when analyzed from the perspective of a human. Yet, who is to say beings beyond our comprehension wouldn't think the same of us?

Unless of course that was your point this entire time and I just stupidly repeated it in a more garrulous manner.


We have society, schooling, media, art, and culture; and every human is unique and individual thanks to thousands of years of these.

Even if we got a fruit fly to speak, it's likely that every other fruit fly will say very similar things as they just have the genetic code but no "culture."

So I think the difference here is caused by society, not by simplicity. Each human carries the combined knowledge of ancestors.


That was exactly my point.

Hart's paper attempts to cover the idea of "sociological" phenomena like this.

I think from an evolutionary biology standpoint, it seems dubious that such a thing could evolve -- even considering a software solution, software that more rapidly propagates will be more common then software that does not.

They might not want to talk to us, but members of that species that actively convert planets into computational substrates would be the more common variants just because on these kinds of timescales they would not have competition.


My favorite solution: consciousness is not that important of a criteria when determining the success of an organism- there are plenty populated planets all over the universe, but none of them care about learning, or exploring the unknown. Our giant brains are just one very specific solution to one very specific problem, and in some environments, there are advantages to not having such a brain.


Sounds like the plot for Spin: https://www.goodreads.com/book/show/910863.Spin


Why software? Seems like a pretty dumb and fragile platform for a quasi-immortal being.


DNA is kinda, like software. We're, what, decades after transistors were invented and people are already talking about implanting chips t smarten us up. Imagine 1 billion years from now. Some civ might have started billions of years before us.


I think that thinking around the Fermi Paradox has the fatal flaw that we think our current technological boom can be extrapolated. We have the feeling that the Kardashev scale, even type 1, is attainable. I'm more technological pessimistic and think it is not attainable. I think the technological boom will peter off, because of hard insolvable questions limited by nature and energy and money. We are enthusiastic now because of past progress and progress we see we can still make. But we will get in diminishing return time and slow stabilization before too long. And progress will become hard until impossible.

Humans will never colonize this galaxy. It's too hard, too expensive, and too pointless. And neither will other intelligent civilizations, of which I guess a small handful at most will exist in the lifetime of our galaxy. Maybe some galaxies do exist where a civilization lives that put everything of their energy on spreading within their galaxies. But even they will not be able to spread to other galaxies. The acceleration of the universe and the extreme distances and nothingness forbids that.


>Humans will never colonize this galaxy. It's too hard, too expensive, and too pointless.

If we stick around long enough, it will eventually, over a long enough scale, become none of those things. We're 95% of the way, as a species, toward the point of spreading out to such a diaspora that no natural force can eradicate us.

Even if we don't end up doing it because of nuclear weapons or mass suicide or some other reason, we would expect at least some other civilizations that reach our level of advancement to reach that last 5% beyond -- unless no other comparable civilizations have ever arisen in the first place.


Why do you think we are in danger of a dramatic slow down in new tech? The tech challenges today are certainly more difficult but our gains compound on themselves. Faster computers allow for previously computationally infeasible techniques, better tech allows for better understanding of biology from which we draw better, biologically inspired techniques, better cultural improvements/ greater access to information allows more people to work on a given problem. I definitely see certain roadblocks like the end of moore's law but there are other ways forward too.


Exponential curves are unnatural. If there is something that constrains growth--such as physical laws about energy constraints--then you'll get a sigmoid curve instead. And it does seem that we're on the slowdown section of that curve for computers. Meanwhile, in biology, we can summarize most of the research results of the past few decades as "it turns out that this is more complex than we thought it was."


> too pointless

It has seemed to me that we _have_ to colonize the galaxy, if we aren't - as a species - going to be forced to resort to periodic mass genocide of exactly the type discussed in the Marvel Studios Infinity War movie.


This is patently false. Hans Rosling and others have explored the statistics of the population growth on Earth, and unless our course of childbirth and death drastically changes, we will cap out at 9-10 billion. There are many factors, but basically, as poorer nations become middle-class they start having fewer than 2 children per couple on average.


The Percolation Theory reasons that Earth might linger in an uninteresting part of the Milky Way: its galactic backwaters. See:

* [1998] http://www.geoffreylandis.com/percolation.htp

* [2014] https://arxiv.org/abs/1404.0204


Oh, okay, so we should be fine until the Vogons come.


Percolation theory [1] is touched on in 3-4 pages in The Eerie Silence: Renewing Our Search for Extraterrestrial Intelligence [2] as well.

[1] https://en.wikipedia.org/wiki/The_Eerie_Silence [2] https://en.wikipedia.org/wiki/Percolation_theory


The notion we'd settle on planetary surfaces makes assumptions on the anthropology of our offspring.

Moving between stars requires us to master living in enclosed habitats for generations. If we do that, why would we bother with planets? We could arrive at the periphery of a system or an accretion disk of a young star, get all the materials to build a couple hundred new habitats and launch them towards the next system.

Would we still want to settle on planetary surfaces?


This is a huge aspect. We've already got lots of precedent of splits between nomadic vs. agricultural societies. I think we would naturally group ourselves into 'travelers' and 'settlers', with a small amount of migration between the two-- and eventually in the long run, species differentiation.


With small populations isolated during the centuries between stops, differentiation will be very fast and could happen within a single longer trip. There are only 10,000 years between wolf and poodle.

In order to avoid it, the groups could start the journey with a set of frozen eggs (or static genomes in whatever medium they find convenient) to be reinserted into the population along the trips, as well as extensive genetic matching between the living populations to ensure a viable long-term gene pool.

And, of course, it's not unreasonable to expect some groups to engage in active eugenics and direct their evolution so they are better suited to their way of live - nomadic or settler.


My “boring” “explanation”: the number of alien intelligences is unknowable and irrelevant because interstellar travel and communication are both near as damnit impossible.


Yeah I wonder if this isn't the simple (and sad) truth. Maybe every intelligent species discovers that the laws of physics are unavoidable hard limits and eventually decide it's a pointless waste of energy to endlessly broadcast signals to other stars and galaxies that are moving, ever faster, further and further away.


Is it just me, or did I not catch any new solutions to the paradox in the article?


They link directly to the new paper:

https://arxiv.org/abs/1902.04450

Snippet from abstract:

> Our steady-state model can constrain the probabilities for an Earth visit by a settling civilization before a given time horizon. These results break the link between Hart's famous "Fact A" (no interstellar visitors on Earth now) and the conclusion that humans must, therefore, be the only technological civilization in the galaxy.


Not just you. The article just summarized some of the more popular theories.


Exactly my thinking. Seems like a whole lot of fluff sans new insight.


It isn't a new solution, it's a contradiction of the paradox. That is, the initial assumptions of the prevalence and proliferation of life are statistically wrong. If we account for x, y, and z, we get the results we observe today.


FTA: "Faster-growing, rapacious societies might peter out before they could touch all the stars."

I don't get this reasoning. The way I see it: every solar system that gets settled is a fresh roll of the dice. Every newly colonized system increases the odds that at least one of these settlements will not self-destruct.

Polynesia is a great analogy. While Easter Island is suffering, Hawaiʻi, Tahiti and hundreds of other islands might still be going strong.


That is if it is somehow economically feasibly for those societies to develop interstellar travel in the first place. When a civilization is occupied with its own struggles and conflicts (the fast-living rapacious ones), they might be incentivized to first exploit local resources before interstellar travel is considered seriously.


Folks, the "Great Filter"[1] is looking us right in the face, in the form of catastrophic climate change[2].

It seems quite a likely hypothesis that any upcoming intelligent species would, like us, ride the horse of fossil fuels right into disaster.

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

[2] http://nymag.com/daily/intelligencer/2017/07/climate-change-...


Climate change won't end life on earth.

It may be an extinction event for humans. Personally I think this is unlikely, but so what if it is? There have been extinction events before, and in fact that is why we are dominating the planet instead of the dinosaurs.


I propose the Morris equation for the odds of alien life finding us:

N = ( R * fp * ne * fl * fi * fc * L ) / ( AI * VR * D * S * M )

Where the Drake equation:

N = The number of civilizations in the Milky Way Galaxy who detect our electromagnetic emmissions

R = Average rate of formation of suitable stars (stars/year) in the Milky Way galaxy

fp = Fraction of stars that form planets

ne = Average number of habitable planets per star

fl = Fraction of habitable planets (ne) where life emerges

fi = Fraction of habitable planets with life where intelligent evolves

fc = Fraction of planets with intelligent life capable of interstellar communication

L = Years a civilization remains detectable

Is divided by the karma equation:

AI = Whether or not aliens have merged with artificial intelligence, become their own God and are no longer interested in us (tends towards 1)

VR = Whether or not aliens have virtual reality akin to the movie The Matrix and live in a hedonistic paradise indefinitely (tends towards 1)

D = Number of psychedelic drugs available to aliens that are at least as compelling as actual reality (tends towards infinity)

S = Percentage of aliens that reproduce sexually or are able to spend a lifetime living in their parents' basement pondering aliens (let's just say 50%)

M = Whether or not aliens use money so contact with extraterrestrials is considered too great a risk (somewhere between 0 and 1)

So as with all things, first contact by extraterrestrials is likely dominated by the need to procreate and money.


Out of millions of species that have appeared on this planet, only 1 made it to space. It could be most aliens never make it to space.


Every single slightly habitable planet in the galaxy could harbor a species that has gone to it's nearest moons and planets, and we wouldn't be able to tell. Every transmission our species has made is indistinguishable from quantum noise at the distance of Alpha Centauri.

In the sense of the Fermi Paradox, zero species on this star have "made it to space".


But one of the tenets of the paradox is so much time has passed you shouldn’t have to work hard to detect alien life, they should have already colonized everywhere.

>Even at the slow pace of currently envisioned interstellar travel, the Milky Way galaxy could be completely traversed in a few million years.


Which assumes that we know what "colonised" looks like.

The entire premise is rather bizarre.

"Let's send out probes that make probes that send out more probes until we've covered the galaxy!"

"But... why would anyone do that?"

It seems obvious until you think about it - and then it stops being obvious.

So you cover your galaxy with probes that keep sending probes back and forth. So what? What have you actually achieved?

If you assume massively superhuman intelligence, what are the odds there are more interesting goals for a civilisation to aim for?


>Which assumes that we know what "colonised" looks like.

100% my problem with this. We're applying a human-centric lens to this. Remember that these theories are based on a life-form that evolved completely separate from anything we have ever known. That doesn't just mean separate body structures, but an entirely novel and completely unknowable circumstance.

This is why these theories all upset me - they should all say things like "we should have seen (an alien species that is similar to us) in this amount of time.

Because fuck only knows what an alien who evolved in a system based on helium or some other gas would think. We certainly can't try to figure it out. The most creative minds in this area are those writing hard sci-fi, I would argue. And most of them only treat that subject with a hand-waving and 'we can never know those mysterious aliens' sort of thing.


Do you expect an alien species that is intelligent enough to create the technology to travel through space will understand math in a way that is at least in principle intelligible to humans?

If yes, is it possible that there are also other ideas that all intelligences are likely to almost necessarily converge on?


The Fermi paradox isn't that there are no aliens, it's that there aren't billions of them, and if there are billions of them not a single one is as expansionist as humanity.


Humanity is not expansionist on a galactic level. Once we are in a position we could make it to the stars, maybe we wouln't do it either, since it might take too much resources that can be spent more efficiently in ensuring our own survival here in this system.


> Humanity is not expansionist on a galactic level

The sun will go boom soon enough, so survival is also temporal. We understand distributed systems are more reliable and require 1+n


>> The sun will go boom soon enough

Depends on your definition of 'soon enough'. The sun will have exhausted its hydrogen supply in about 5 billion years. In about 1 billion years, the temperature on this planet will probably become unsustainable for life.

Your right of course, but that event is so far in the future, I doubt we can speak of anything even resembling humanity by that time. I was thinking in timeframes of about a ten to hundred thousand years, not billions.


But a civilization doesn't have to 'aim for' that. If you can create the first probe, it does the rest. The right question is, what are the chances that nobody in any civilization ever, thought to create that first probe?

If we can think of it (and here we are, thinking of it) then why wouldn't any other civilization contain a member that thinks of it? Because humans are the greatest smartest species in the galaxy? Now that would be a far-fetched idea.



Although compared to reaching space it's relatively simple to prevent other species from ever reaching space, as the Homo Sapiens may well have done.


"Out of millions of species that have appeared on this planet, only 1 made it to space."

So far. In another ten million years, several more might make it.


On a related note, this recent article explores the idea that life may be inevitable. (An idea going back at least as far as Ilya Prigogine.)

https://qz.com/1539551/is-the-universe-pro-life-the-fermi-pa...

"According to the inevitable life theory, biological systems spontaneously emerge because they more efficiently disperse, or “dissipate” energy, thereby increasing the entropy of the surroundings. In other words, life is thermodynamically favorable."


Still takes a hell lot of time to develop to a point of spreading from one star to another.

Isaac Arthur has a whole series of podcasts that cover all the current ideas around the Fermi paradox, which ones are more or less likely and why. Really interesting stuff. Here's a link to his Fermi Paradox playlist on Youtube: https://www.youtube.com/playlist?list=PLIIOUpOge0LulClL2dHXh...

He's also got a Soundcloud for audio-only format and a Facebook group.


The problem with the Fermi paradox is that it requires you to assume the existence of spacefaring civilizations, with all the expected technology. The possibilities are just to many, and we only know relatively well our own solar system, we don't know how normal it is (recent observations suggests it could be at least kind of odd) or how normal Earth is. As it is quoted, Earth for us is all there is and it would seem completely normal to us even if for any well educated alien it could look obviously artificial.

The Black planet hypothesis: There are planets intentionally cut off the bigger society. Spacefaring civilizations are largely immortal, and would be kept in an evolutionary stasis, so there are planets that are intentionally kept isolated and prisoners are kept on them in a pre galactic stage as a way to keep evolving.

Impending doom hypothesis: An immense catastrophe is expected to happen in the general area, and all the surrounding space has been evacuated.

Presumed recluse hypothesis: Our solar system is recognized as obviously artificial, yet attempts to communicate are met with no response. We are perceived as unwilling to communicate and intentionally left alone.

Planets are primitive hypothesis: More advanced civilizations have no interest in settling planets, planets are seen as primitive as living in a rainforest is to us, while advanced civilizations normally live in artificial habitats.

Unpleasant star hypothesis: The solar wind, neutrino flow or other forms of radiation may be disruptive to highly advanced technology, so the sources are avoided unless necessary.


> it requires you to assume the existence of spacefaring civilizations

That's one of the variables in the Fermi equation. Set it to zero, and the paradox vanishes.


Well, yeah. "with all the expected technology" is the important part.


That's another variable.


I'm not sure I understand what you mean. Do you disagree with me on something?


The Fermi Paradox requires you to assume a lot of things that we just don't know. They're all variables; you just named two of the more interesting and contentious variables out there.

Until recently, even the average number of planets around a given star was really unknown; arguably, it's still an incredibly fuzzy figure, one that changes constantly as our telescopes and observation methods improve.

There's no one "THE" problem with the Fermi Paradox.


It is "the" problem, as it makes it impossible to tell what is natural and what is intentionally created by advanced civilizations, making it impossible to know if the paradox even exists at all, since we may be looking at "obvious" signs of alien activity, yet be completely incapable of recognizing them as such.

> The Fermi Paradox requires you to assume a lot of things that we just don't know.

Dont agree on* There is an averge number of planets. The problem some people have is in regard to sample size.


tl;dr it shouldn't take as long as previously thought to colonize the galaxy based on a simulation.

That's largely an irrelevant finding that adds nothing to the Fermi Paradox because the time it would take to colonize the galaxy is so small in cosmological terms that shortening it is irrelevant.

What's more the authors decided to modeled the suitability of star systems impeding colonization when this seems like it would be largely irrelevant because starfaring civilizations seem highly unlikely to be planet-bound as this is a pretty inefficient use of mass. The Dyson Sphere (Swarm) seems much more likely at which point the only thing of value is the star. The planets are just piles of raw materials ultimately.

Obligatory plug for Isaac Arthur who has a series of videos on the Fermi Paradox. Here's an early one [1]. There are 10-20 others that go into particular aspects.

[1] https://www.youtube.com/watch?v=rDPj5zI66LA


i dont know why people think there is a paradox. the way that life springs from barren rock is not known. if we dont know how that works, then we cant assign a probability to it happening on a given planet. people just assume that the probability is very high. it could be next to nothing for all we know, small enough so that even the entire universe only produces one. the paradox is all based on huge assumptions. there is no paradox until we prove that the probability is high.

well thats not true, because you need life but you also need intelligence. and again, everyone assumes that if you have life it will eventually become intelligent. and people assume that if life is intelligent it will eventually build space shuttles. its all a huge, huge assumption. look at all the animals that qualify as intelligent. some birds and monkeys are hugely intelligent, but they dont build space shuttles. this shows that intelligence doesnt equal space shuttles and that even when life springs up, and even when it becomes intelligent, it still could be super unlikely that it will build space shuttles.

hell, there are even humans that might have never built space shuttles. there are indigenous communities all over the planet that never developed technology and probably never would have. when you live in a warm climate and food is abundant, there may never be a reason to.

it is unproven that it is likely at all for space-shuttle level intelligence to spring up from bare earth. there is no paradox. its probably just really unlikely.


If we assume that most stars are in a sort of orbit around the core, wouldn't the stars stay in relatively the same place? And regardless, wouldn't they still require the ability and timing to take advantage of those moments the solar system's get close to each-other? That still requires a monstrous amount of technology and energy, doesn't it?


An lower orbit is faster than a higher orbit (orbits trade speed for height a bit like pendulums). So on long enough timescales your neighbors change as every solar system is on a slightly different orbit.

> That still requires a monstrous amount of technology and energy, doesn't it?

If we wanted we could spend ~10 years R&D on a practical nuclear pulse drive and send a probe at 10% the speed of light to the nearest solar system. That makes 40 years travel time for the probe and 4 years communication delay. Add some time for acceleration , deceleration and exploration and we might be able to send a colony ship 60 years after the probe left. Another 40 years travel time and we have a colony on another solar system 110 years from now, at a price point attainable for a large corporation or a medium sized country (after all most of the cost is research spread over half a century)

Now imagine what a more advanced civilization with more incentives for colonisation could do.


We could be mining asteroids today for materials that are scarce on earth, yet we don't. There isn't enough incentive to drive anyone to do that yet. What makes us so sure advanced civilizations recognize a worthwhile incentive to colonize other worlds around other stars?


Yeah, even in 2019 here on Planet Earth, some of our limitations on space exploration are self-imposed. If we were more willing to use nuclear power in our rockets, we would probably be much further along with having stuff in space.

Presumably all those aliens that aren't getting to space because of their nuclear wars are that much more likely to be willing to use nuclear power for space travel, too, in which you'd think at least one of them would have threaded the needle between nuking their home planet, but getting to self-sustaining in space, first.


Have you considered the acceleration that such a nuclear pulse drive creates? I am totally with you on the topic, not just with living organisms. If it were only electronics that need to be transported the force by acceleration wouldn't be such a huge issue. But any living organism that is to be transported is such a way needs to be able to adapt to pretty extreme acceleration/deceleration forces.


Maybe parent is a bit optimistic, both in terms of costs and travel time, but the orders of magnitude are approximately correct. Project Orion did come up with theoretical designs that were feasible for human travel (1 g acceleration).

https://en.wikipedia.org/wiki/Project_Orion_(nuclear_propuls...


Huh, interesting... I always thought this kind of propulsion would be an almost instantaneous transmission of energy, thus enormous acceleration in a small timeframe. Thanks for the link...


In this scenario who is "we" that invests that time and energy?


I think The Expanse had the right idea with a powerful religious organisation like the Mormons. I could also imagine a totalitarian regime doing it to unite their people under a common goal (similar to the space race being a source of national pride especially for the Soviets). Or maybe a billionaire chooses to dedicate his money to such an endeavour (a Bill Gates who loves space as much as Elon Musk).

I agree it's unlikely that our civilization does it now. But given some time and a handful of civilizations and it seems like it should happen all the time (even more so if overpopulation or collapsing ecosystems were serious problems for some civilisations)


Stars closer to the core move faster, so it's more like a whirlpool than a rotating disk.


Do we have an estimate for likelihood of noticing earth as a life containing planet as a function of how far away we might imagine ourselves in the galaxy given current methods?


I remember reading than from a radio emissions standpoint, we wouldn't detect our activity if it was happening on a planet in the closest star system. I don't know about other methods but very interested too.


> It’s possible that the Milky Way is partially settled, or intermittently so; maybe explorers visited us in the past, but we don’t remember, and they died out. The solar system may well be amid other settled systems; it’s just been unvisited for millions of years.

This, however, does not explain the complete lack of artifacts or observable technological signatures.


I like Stargate SG-1's solution to this: Evidence is all around us (both in mythology and ancient artifacts), we just don't recognize it.


Oh, but don't forget that one congregation of dust particles that flew by us really fast, which _might_ (who am I kidding - we all know it's 100%!) have been an alien probe.

Oumuamua is what it's been called.


An alternate explanation for that is that it was effectively a giant fractal dust bunny left over from the formation of a solar system. The fractal structure explains why it is so light. As for how fast it is traveling, it was traveling at close to the average speed of stars around us. The 25 km/s that we saw is mostly how fast we are traveling, not it. (The Sun bounces up and down through the galactic plane. We are currently "above" it and heading up. We'll next arrive back where we are in about 30 million years.)

There might be a lot of objects like that out there. In fact the fact that we randomly saw one suggests that there are many orders of magnitude more of them than we had realized.


The PBS Space Time video series, written and hosted by Lehman College astrophysicist Matthew O’Dowd, has a nice analysis of popular 'Oumuamua theories:

"'Oumuamua Is Not Aliens" https://www.youtube.com/watch?v=wICOlaQOpM0


The obvious, but hard to swallow solution to the Fermi paradox: aliens are a product of science-fiction, and don't exist. Not everything we can come up with in thought experiments has a real life counterpart.


But unlike time travel or flux capacitors aliens do have a real life counter part, us. We know life exists, it's not a stretch to imagine there is more of it.


Before we can postulate a realistic generalized conclusion, we need a sample size larger than 1. Right now the scientific, rational position on Fermi's paradox is that outside of Earth the universe is completely devoid of life.


That is one rational position, not THE rational position. I'd say it is the only empiricist position to take, but not all rational arguments are necessarily empiricist.


In 1980 was the scientific, rational position that outside of our own Solar System there were no more planets in the universe? I have a hard time calling that a rational position despite the lack of observed exoplanets at that date.


Obviously the aliens wait to make contact until there is only one species left on the planet.

We're almost there!


I don't think the tardigrades are going to be able to respond.


There are lots of aliens. Heck, there were aliens attending certain underground raves in Seattle in the 90's.

The Fermi Paradox is about the psychology of humans, not the external universe.

As soon as you can psychologically handle communication with other folks it becomes easy to meet and greet them. There are "outreach coordinators" who volunteer to take humans on jaunts within the local neighborhood (nearby star systems, etc.)

Our planet experienced a disaster ("dis": separation; "aster": stars; "disaster" is a separation from the stars) a few thousand years ago (Younger Dryas comet) and we've been recovering ever since. Our neighbors have been helping us for all that time. Psychologically our species is an accident victim and we have been recovering from the trauma, which is why our history is so messed up and violent. Normal sentient beings are mensches. We'll get there.


I guess I just don’t understand this kind of comment. Are you being serious? Can you cite any sort of scientifically verified evidence for these claims? It sounds nice and all, but if you’re legitimately being serious, how are we ever supposed to believe this without hard evidence?


> Are you being serious?

Yes. (Although by bringing this up in this thread I could be accused of trolling, and it would be hard to defend myself. Sorry.)

> Can you cite any sort of scientifically verified evidence for these claims?

It depends on what you accept as "scientifically verified evidence". My whole point is the Fermi Paradox is about human epistemology and belief systems, not the actual contents of human experience (which include UFOs and little green men as well as lots of other weird shit.)

> It sounds nice and all, but if you’re legitimately being serious, how are we ever supposed to believe this without hard evidence?

You're not, evidently. Honestly, I'd settle for folks just getting over their personal problems and returning to a harmonious lifestyle, and if I knew how to do that I would. A lot of the aliens and other people around here are working on the problem, but we're a tough nut to crack and time is running short.


I’d love it if you could unpack some of this. You gloss over some pretty big statements here.


I need more of this as well. Comments, books... whatever.

Even if it isn't real its a fascinating and beautiful story.


I'll try (and yes, some of this is allegory, or metaphor, or whatever.)

Normal human consciousness is very different from what we have now. Normally we're very hip and mellow and happy. (Cf. G. I. Gurdjieff)

The dimly remembered golden age was just normal life, then there was this accident (maybe it was the still-hypothetical Younger Dryas Comet Impact, maybe not.) There was a global disaster (flood, the oldest human societies have actual oral traditions handed down from the survivors.) In the aftermath of that we fell into a kind of Lord of the Flies scenario and developed agriculture and cities (both pathological aspect of our sundering from Nature/Reality.)

All of what we think of as human history is the story of accident victims recovering from a terrible disaster, rising up out of a kind of nightmare or coma, and returning to normal healthy existence, which feels so good (in comparison to the trauma-filled previous existence) that e.g. Christians call it the Kingdom of Heaven.

As we heal our individual and collective trauma we have rediscovered e.g. the basic structure of the Universe (astronomy, cosmology, the Standard Model) and begun to live in harmony with Nature (e.g. Permaculture and other forms of applied ecology). As a side effect of living in harmony with Nature we can communicate with other intelligent beings around us.

https://en.wikipedia.org/wiki/The_Secret_Life_of_Plants (esp. Findhorn)

https://ase.tufts.edu/biology/labs/levin/




Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: