Grabby aliens are advanced civs who change the stuff they touch in big visible ways, and who keep expanding fast until they meet each other. Our recent analysis suggests that they appear at random stars roughly once per million galaxies, and then expand at roughly half the speed of light. Right now, they have filled roughly half of the universe, and if we join them we’ll meet them in roughly a billion years. There may be far more quiet than grabby alien civs out there, but those don’t usually do much or last long, and even the ruins of the nearest are quite far away.
If grabby civ expansion is driven by acquisition of mass/energy resources, and assuming those are roughly equally available in all directions, once such a civ is more than a few thousand years old, it occupies a rough sphere.
The interior of the sphere is stuck with Malthusian limits - resources there are scarce, lots of transactions are zero-sum, and expansion doesn't help because the edge of the sphere is too far away to matter. (Maybe they even lose the ability for interstellar travel, if that's expensive in resource terms.)
At the edges, expansion and acquisition of new resources is possible so they live in what you've called a "dream time", temporarily escaping Malthusian limits.
Living conditions, and therefore culture, are probably extremely different at the edges vs. the interior.
Expansion continues, enlarging the sphere, until they meet a neighboring grabby civ.
At that point, every contact with the neighboring civ is a first-contact, because the spheres are large relative to the speed of light. So the contact surface expands far faster than the speed of light.
If every contact is a first contact, I wonder how much military coordination and planning re contact is possible at all. Esp. if the interior civ is very different from the edge civ. And movement of resources over long distances (any significant fraction of a civ's diameter) is pointless due to the time required.
Thanks for your response to my comment. Yes, software bugs and and unforseen consequences seem like a far more likely source of such diverging factions. Especially if the "civilisation" involved is highly complex.
This problem has been nagging me too ever since I pinged you prior to your release of the Grabby Aliens paper. I of course feel that there is an unfortunate ongoing power dynamic between "intelligence" qua strength in conflict scenarios versus the attainment of supererogatory pleasures. In my own lifestyle documentation I aspire towards striking that balance with optimised entertainment as well as the production of intelligence. But for our broader species in maximising expected aggregate utility across the lightcone we run into potential problems agentically optimising aggressively towards strength over the actual realisation of the sorts of qualiae we deem to be the valuable good kind. The tension between the Survival and the Flourishing.
The defense capabilities of space are so strong that wars are nearly impossible. Any civilization that holds a star system can defend it from intruders. By converting planets and asteroids into fortresses, small invading fleets can be detected and eliminated at long distances. Antimatter kinetic near-light weapons can cause destruction of planets, but not of starships in the local Oort cloud.
Nevertheless, attacks on gas giants by antimatter weapons could hypothetically cause a short-lived explosion, releasing about 1000 years of solar luminosity in a few seconds, which is much less than a nova or supernova (mostly from lithium-6 burning in some deep layer). This would result in the destruction of most starships and the filling of a star system with radioactive debris, inhibiting space travel. Yet, some bases could survive in larger asteroids and in the outer Oort cloud. Such explosions could act as signs of space battles.
Actually it's not that hard to make a storage device safe from data corruption. Just add enough checksums and error correcting codes and redundant drives, and any random data corruption can be identified and fixed before it causes problems. Each additional piece of error-correcting hardware gives an exponential reduction in uncorrected error rates. If a future civilization really cares about preventing mutated and hostile factions, it could easily devote the extra resources to make the chance of such mutation so low it will virtually never happen despite the time scales involved.
A harder to crack issue that could create internal conflicts would be software bugs with unintended consequences.
A moat? You propose burning all bridges in a sphere? That would limit your own access to resources and allow your territory to be completely surrounded. It seems like a recipe for disaster to me.
There are some natural moats - around galaxies, for example. IMO, it is not clear what the dynamics of life spreading between galaxies is. Fortunately, we don't need to worry so much about that for a little while.
Invasions would be most likely to succeed if there's a big difference in tech levels or available resources. A microorganism attacking a much larger colony of similar microorganisms seems like a misleading metaphor for this situation. You could possibly argue that differing tech levels are unlikely since civs relatively rapidly reach tech maturity and run into physical limits. That's a hypothesis - though probably a mistaken one, IMO.
I think grabby civs will be far from what we would call a civilisation made up of people living on planets. They would be a bunch of intelligent systems and robots probably pursuing the aquisition of resources in order to gain domiance over other (internal) parties that do the same but not in the way they like. That's how one might get conflicts.
Disagreement might stem from mutations in their instructions. No storage medium is safe from corruption, and given the enormous timescales in which these robots might operate, errors in replication might lead to new factions that don't act in the way they were intended to do.
Or they might mutate to steal (i.e. hack into) already build systems for their own use. Kind of analogous to the way some lifeforms on Earth evolved to just eat others and thereby profiting off of the energy contained in tnutrients that the victim broke down for itself.
At least that would be an argument in favor of internal conflict being more common.
I think grabby civs will be far from what we would call a civilisation made up of people living on planets. They would be a bunch of intelligent systems and robots probably pursuing the aquisition of resources in order to gain domiance over other (internal) parties that do the same but not in the way they like. That's how one might get conflicts.
Disagreement might stem from mutations in their instructions. No storage medium is safe from corruption, and given the enormous timescales in which these robots might operate, errors in replication might lead to new factions that don't act in the way they were intended to do.
Or they might mutate to steal (i.e. hack into) already build systems for their own use. Kind of analogous to the way some lifeforms on Earth evolved to just eat others and thereby profiting off of the energy contained in the nutrients that the victim broke down for itself.
At least that would be an argument in favor of internal conflict being more common.
I'm imagining a castle with a moat that's at minimum of 4.5 light years wide, and possibly orders of magnitude wider. You have a point about destruction being easier than construction, but unless you're interested in simple terrorism, the destruction you can do across light years won't be enough to make an inhabited system ready for you to move in. I disagree that you can break enough local stuff to bomb a civilization back to the stone age, not if every star is orbited by a swarm of billions of orbiting computers like I would expect it to be. Sure, you can smash up some of them, but the locals will just put the pieces back together, and you gain nothing. Like I said above, if you're mold, you don't have much to gain from invading an already-moldy orange.
Re: "Would-be attackers need to bring lots of stuff, and the defenders already have command over the local stuff." Home turf advantage has also hindered invasions on Earth - but it hasn't completely prevented them. Near to the border, invaders don't have to carry their stuff very far. There's also the advantage of destruction over construction to consider. Control over local stuff is quite easy to break with few resources by bombing everything back into the stone age - as you mention. Then, the sides of the local conflict become more evently matched.
The way I picture it, there would be little difference between colonization and war when it comes to encounters of advanced grabby civs. Basically, these civs just develop a routine for assimilating newly-reached systems. Let's say that routine is to disassemble stars, planets, and asteroids, turn suitable elements computronium, and light elements into fusion fuel. That's generally done by shooting very fast self-unpacking seeds that take root in virgin territory and infest the local resources, like a mold infests an orange. Once a system becomes our computronium, it's easy to deny somebody else's seeds from taking root there; those seeds are tiny and fragile. Interesting might be border systems where both we and another civ get a foothold. That could make for a pretty intense in-system wars. Still, most systems will be undisputed, and I think they might be pretty easy to defend once sufficiently infested. Would-be attackers need to bring lots of stuff, and the defenders already have command over the local stuff. And yeah, successful invaders would get to break down our computronium and turn it into their computronium, but at immense effort. Moldy oranges are generally ignored by other molds.
So I'm imagining pretty stable frontiers at the bubble boundaries, which might look like a de-facto truce. This would end only if our computronium could somehow be cheaply ruined from a distance. Hacking is one option. If our computronium has an exploitable vulnerability, the hack might spread faster than our "We've been hacked and urgently need to patch system xyz" message. That would be a great way to eat up a civilization at almost c. A second invasion strategy would be to just burn down our border systems by pelting us with small black holes that consume and grow from our computronium, kind of sterilizing that system by speeding up its descent into entropy. The aliens would gain nothing from a "win" like that, except maybe some deterrence and the knowledge that the frontier between us and them is one sterile system wider.
What's hard for me to picture is that aliens would assemble a massive fleet and send it to conquer the Milky Way, which is by then made up of our computronium. Whatever they could send would be puny compared to the Milky Way itself.
I suggest it would be good to start these grabby alien posts with an epistemic status, like Scott Alexander does, and acknowledge that all the assumptions have incredibly high uncertainty with orders of magnitude variance.
If grabby civ expansion is driven by acquisition of mass/energy resources, and assuming those are roughly equally available in all directions, once such a civ is more than a few thousand years old, it occupies a rough sphere.
The interior of the sphere is stuck with Malthusian limits - resources there are scarce, lots of transactions are zero-sum, and expansion doesn't help because the edge of the sphere is too far away to matter. (Maybe they even lose the ability for interstellar travel, if that's expensive in resource terms.)
At the edges, expansion and acquisition of new resources is possible so they live in what you've called a "dream time", temporarily escaping Malthusian limits.
Living conditions, and therefore culture, are probably extremely different at the edges vs. the interior.
Expansion continues, enlarging the sphere, until they meet a neighboring grabby civ.
At that point, every contact with the neighboring civ is a first-contact, because the spheres are large relative to the speed of light. So the contact surface expands far faster than the speed of light.
If every contact is a first contact, I wonder how much military coordination and planning re contact is possible at all. Esp. if the interior civ is very different from the edge civ. And movement of resources over long distances (any significant fraction of a civ's diameter) is pointless due to the time required.
Thanks for your response to my comment. Yes, software bugs and and unforseen consequences seem like a far more likely source of such diverging factions. Especially if the "civilisation" involved is highly complex.
This problem has been nagging me too ever since I pinged you prior to your release of the Grabby Aliens paper. I of course feel that there is an unfortunate ongoing power dynamic between "intelligence" qua strength in conflict scenarios versus the attainment of supererogatory pleasures. In my own lifestyle documentation I aspire towards striking that balance with optimised entertainment as well as the production of intelligence. But for our broader species in maximising expected aggregate utility across the lightcone we run into potential problems agentically optimising aggressively towards strength over the actual realisation of the sorts of qualiae we deem to be the valuable good kind. The tension between the Survival and the Flourishing.
The defense capabilities of space are so strong that wars are nearly impossible. Any civilization that holds a star system can defend it from intruders. By converting planets and asteroids into fortresses, small invading fleets can be detected and eliminated at long distances. Antimatter kinetic near-light weapons can cause destruction of planets, but not of starships in the local Oort cloud.
Nevertheless, attacks on gas giants by antimatter weapons could hypothetically cause a short-lived explosion, releasing about 1000 years of solar luminosity in a few seconds, which is much less than a nova or supernova (mostly from lithium-6 burning in some deep layer). This would result in the destruction of most starships and the filling of a star system with radioactive debris, inhibiting space travel. Yet, some bases could survive in larger asteroids and in the outer Oort cloud. Such explosions could act as signs of space battles.
Actually it's not that hard to make a storage device safe from data corruption. Just add enough checksums and error correcting codes and redundant drives, and any random data corruption can be identified and fixed before it causes problems. Each additional piece of error-correcting hardware gives an exponential reduction in uncorrected error rates. If a future civilization really cares about preventing mutated and hostile factions, it could easily devote the extra resources to make the chance of such mutation so low it will virtually never happen despite the time scales involved.
A harder to crack issue that could create internal conflicts would be software bugs with unintended consequences.
A moat? You propose burning all bridges in a sphere? That would limit your own access to resources and allow your territory to be completely surrounded. It seems like a recipe for disaster to me.
There are some natural moats - around galaxies, for example. IMO, it is not clear what the dynamics of life spreading between galaxies is. Fortunately, we don't need to worry so much about that for a little while.
Invasions would be most likely to succeed if there's a big difference in tech levels or available resources. A microorganism attacking a much larger colony of similar microorganisms seems like a misleading metaphor for this situation. You could possibly argue that differing tech levels are unlikely since civs relatively rapidly reach tech maturity and run into physical limits. That's a hypothesis - though probably a mistaken one, IMO.
I think grabby civs will be far from what we would call a civilisation made up of people living on planets. They would be a bunch of intelligent systems and robots probably pursuing the aquisition of resources in order to gain domiance over other (internal) parties that do the same but not in the way they like. That's how one might get conflicts.
Disagreement might stem from mutations in their instructions. No storage medium is safe from corruption, and given the enormous timescales in which these robots might operate, errors in replication might lead to new factions that don't act in the way they were intended to do.
Or they might mutate to steal (i.e. hack into) already build systems for their own use. Kind of analogous to the way some lifeforms on Earth evolved to just eat others and thereby profiting off of the energy contained in tnutrients that the victim broke down for itself.
At least that would be an argument in favor of internal conflict being more common.
I think grabby civs will be far from what we would call a civilisation made up of people living on planets. They would be a bunch of intelligent systems and robots probably pursuing the aquisition of resources in order to gain domiance over other (internal) parties that do the same but not in the way they like. That's how one might get conflicts.
Disagreement might stem from mutations in their instructions. No storage medium is safe from corruption, and given the enormous timescales in which these robots might operate, errors in replication might lead to new factions that don't act in the way they were intended to do.
Or they might mutate to steal (i.e. hack into) already build systems for their own use. Kind of analogous to the way some lifeforms on Earth evolved to just eat others and thereby profiting off of the energy contained in the nutrients that the victim broke down for itself.
At least that would be an argument in favor of internal conflict being more common.
I'm imagining a castle with a moat that's at minimum of 4.5 light years wide, and possibly orders of magnitude wider. You have a point about destruction being easier than construction, but unless you're interested in simple terrorism, the destruction you can do across light years won't be enough to make an inhabited system ready for you to move in. I disagree that you can break enough local stuff to bomb a civilization back to the stone age, not if every star is orbited by a swarm of billions of orbiting computers like I would expect it to be. Sure, you can smash up some of them, but the locals will just put the pieces back together, and you gain nothing. Like I said above, if you're mold, you don't have much to gain from invading an already-moldy orange.
Isn't that just a cult membership signal? I don't see most people doing this. Much more practical, useful and common would be an abstract.
Re: "Would-be attackers need to bring lots of stuff, and the defenders already have command over the local stuff." Home turf advantage has also hindered invasions on Earth - but it hasn't completely prevented them. Near to the border, invaders don't have to carry their stuff very far. There's also the advantage of destruction over construction to consider. Control over local stuff is quite easy to break with few resources by bombing everything back into the stone age - as you mention. Then, the sides of the local conflict become more evently matched.
The way I picture it, there would be little difference between colonization and war when it comes to encounters of advanced grabby civs. Basically, these civs just develop a routine for assimilating newly-reached systems. Let's say that routine is to disassemble stars, planets, and asteroids, turn suitable elements computronium, and light elements into fusion fuel. That's generally done by shooting very fast self-unpacking seeds that take root in virgin territory and infest the local resources, like a mold infests an orange. Once a system becomes our computronium, it's easy to deny somebody else's seeds from taking root there; those seeds are tiny and fragile. Interesting might be border systems where both we and another civ get a foothold. That could make for a pretty intense in-system wars. Still, most systems will be undisputed, and I think they might be pretty easy to defend once sufficiently infested. Would-be attackers need to bring lots of stuff, and the defenders already have command over the local stuff. And yeah, successful invaders would get to break down our computronium and turn it into their computronium, but at immense effort. Moldy oranges are generally ignored by other molds.
So I'm imagining pretty stable frontiers at the bubble boundaries, which might look like a de-facto truce. This would end only if our computronium could somehow be cheaply ruined from a distance. Hacking is one option. If our computronium has an exploitable vulnerability, the hack might spread faster than our "We've been hacked and urgently need to patch system xyz" message. That would be a great way to eat up a civilization at almost c. A second invasion strategy would be to just burn down our border systems by pelting us with small black holes that consume and grow from our computronium, kind of sterilizing that system by speeding up its descent into entropy. The aliens would gain nothing from a "win" like that, except maybe some deterrence and the knowledge that the frontier between us and them is one sterile system wider.
What's hard for me to picture is that aliens would assemble a massive fleet and send it to conquer the Milky Way, which is by then made up of our computronium. Whatever they could send would be puny compared to the Milky Way itself.
I suggest it would be good to start these grabby alien posts with an epistemic status, like Scott Alexander does, and acknowledge that all the assumptions have incredibly high uncertainty with orders of magnitude variance.
Radius.
Does half the speed of light refer to the rate of change of a civ's sphere's radius or its volume?
You might be very confident that both of you knew you had a lot more resources to fight with than did they.