7 ms·

There are things like paperclip maximizers (http://wiki.lesswrong.com/wiki/Paperclip_maximizer http://wiki.lesswrong.com/wiki/Paperclip_maximizer). Even a regul

There are things like paperclip maximizers (http://wiki.lesswrong.com/wiki/Paperclip_maximizer http://wiki.lesswrong.com/wiki/Paperclip_maximizer). Even a regular alien civilization is likely to be very, very different than us in entirely unpredictable ways (http://lesswrong.com/lw/so/humans_in_funny_suits/ http://lesswrong.com/lw/so/humans_in_funny_suits/). They may not have a moral system like ours, they may not value our lives at all. Such a civilization might destroy us simply because they don't want us to become too advance and compete with them. Or possibly become a threat to them. Or maybe just because they want our solar system someday. I'm not claiming that any of this is certainty. It really is very difficult to predict. But I caution against anthropomorphic reasoning. I also think it would be wise to be cautious about taking these kinds of risks, if we can.

saulrh 13y ago
There's also the "Killing Star" argument[1]. Basically, when your weapons are relativistic planet-crackers, any civilization you can observe is an existential threat: any civilization that shoots destroys another [2]. Civilizations have survived evolution and presumably other first contacts, and will probably be willing to shoot first to avoid being shot. Any civilization you observe will go through this thought process, realize you are going through the same thought process, and will proceed to shoot because they assume you will shoot. [1] http://www.projectrho.com/public_html/rocket/aliens.php#id--The_Fermi_Paradox--The_Killing_Star http://www.projectrho.com/public_html/rocket/aliens.php#id--... [2] http://www.projectrho.com/public_html/rocket/spacegunexotic.php#id--Relativistic_Weapons--The_Killing_Star http://www.projectrho.com/public_html/rocket/spacegunexotic....
- darkmighty 13y ago
  Too simplistic reasoning imo. If your civ can build impossibly large relativistic weapons, can't it survive an impact of one? Wouldn't the huge cost of building those/providing energy favor the pacifists? And so on. I think it's essentially impossible to predict what to expect from a point of view of specific technologies, and a more fundamental approach is more interesting: given the possible gains we might have, and the possible threats (termination should have a fair likelihood), the expected utility from contact should be hugely skewed negatively -- assuming -inf utility for termination and +finite for whatever gains.
  - Houshalter 13y ago
    At that level of technology, destroying primitive societies like ours would be trivial. It wouldn't necessarily take an impossibly large relativistic weapon. Maybe just a lot of nukes, carefully designed super-virus that wipes out our species, or sending a single self-replicating nanobot (or superior life form) to our planet. We don't know what other destructive technologies are possible. >the expected utility from contact should be hugely skewed negatively -- assuming -inf utility for termination and +finite for whatever gains. If you are assuming death is -inf, almost any action has infinite negative utility. Also see pascal's mugging (http://lesswrong.com/lw/kd/pascals_mugging_tiny_probabilities_of_vast/ http://lesswrong.com/lw/kd/pascals_mugging_tiny_probabilitie...).
    - akiselev 13y ago
      >> Maybe just a lot of nukes, carefully designed super-virus that wipes out our species, or sending a single self-replicating nanobot The insane energy scales of relativistic travel (especially between .9 and .9999-repeating) combined with the ridiculously advanced technologies throws any speculation out of the window. Interstellar nukes surviving relativistic deep space travel while guiding themselves to a solar systems light years away (and with your ability to turn measure in tiny radians of the same)? Super-virus that operates across massive evolutionary scales while, again, surviving said travel (and for kicks only 240 of the known 500 amino acids are naturally occurring [1] and it depends entirely on random mutations and environment)? Self-replicating nanobots, one half of the holy grail (billions of years being the other) that would place in the hands of mortals that which allowed natural selection to give birth to the aforementioned aliens? At the technological level of relativistic weapon systems + nanobots you're basically talking the ability to do whatever you wanted. If they can have near omnipotent self-replicating nanobots flying to destroy civilizations, why not to slightly shift their stellar orbit by consuming the rest of the solar system (relativistic guidance and navigation is an entirely different ordeal) to add some interstellar non-determenacy or send out nanobots to replicate their civilization on hospitable solar systems? Sending out a bunch of relativistic nukes is probably going to be one hell of an expense (don't imagine you'd be able to do a gravitational sling shot without a black hole or perfect premonition). If the civilization was that scared of being wiped out, it'd probably be cheaper to build a single defense for their civilization rather than a bunch of attack platforms intended to wipe out other civilizations.
hnha 13y ago
or they could be the complete opposite, sad about our infighting and hate between our own. they might come to liberate us so we can live happier lives.
- Houshalter 13y ago
  That's very true. Although I think there is a good chance they could have incompatible values to us. It'd be pretty unlikely for two intelligent beings to evolve exactly the same morality. Imagine if they killed us all for eating meat, or if they were strong environmentalists like in The Day the Earth Stood Still. Or something else completely alien to us.

saulrh 13y ago

There's also the "Killing Star" argument[1]. Basically, when your weapons are relativistic planet-crackers, any civilization you can observe is an existential threat: any civilization that shoots destroys another [2]. Civilizations have survived evolution and presumably other first contacts, and will probably be willing to shoot first to avoid being shot. Any civilization you observe will go through this thought process, realize you are going through the same thought process, and will proceed to shoot because they assume you will shoot. [1] http://www.projectrho.com/public_html/rocket/aliens.php#id--The_Fermi_Paradox--The_Killing_Star http://www.projectrho.com/public_html/rocket/aliens.php#id--... [2] http://www.projectrho.com/public_html/rocket/spacegunexotic.php#id--Relativistic_Weapons--The_Killing_Star http://www.projectrho.com/public_html/rocket/spacegunexotic....

darkmighty 13y ago
Too simplistic reasoning imo. If your civ can build impossibly large relativistic weapons, can't it survive an impact of one? Wouldn't the huge cost of building those/providing energy favor the pacifists? And so on. I think it's essentially impossible to predict what to expect from a point of view of specific technologies, and a more fundamental approach is more interesting: given the possible gains we might have, and the possible threats (termination should have a fair likelihood), the expected utility from contact should be hugely skewed negatively -- assuming -inf utility for termination and +finite for whatever gains.
- Houshalter 13y ago
  At that level of technology, destroying primitive societies like ours would be trivial. It wouldn't necessarily take an impossibly large relativistic weapon. Maybe just a lot of nukes, carefully designed super-virus that wipes out our species, or sending a single self-replicating nanobot (or superior life form) to our planet. We don't know what other destructive technologies are possible. >the expected utility from contact should be hugely skewed negatively -- assuming -inf utility for termination and +finite for whatever gains. If you are assuming death is -inf, almost any action has infinite negative utility. Also see pascal's mugging (http://lesswrong.com/lw/kd/pascals_mugging_tiny_probabilities_of_vast/ http://lesswrong.com/lw/kd/pascals_mugging_tiny_probabilitie...).
  - akiselev 13y ago
    >> Maybe just a lot of nukes, carefully designed super-virus that wipes out our species, or sending a single self-replicating nanobot The insane energy scales of relativistic travel (especially between .9 and .9999-repeating) combined with the ridiculously advanced technologies throws any speculation out of the window. Interstellar nukes surviving relativistic deep space travel while guiding themselves to a solar systems light years away (and with your ability to turn measure in tiny radians of the same)? Super-virus that operates across massive evolutionary scales while, again, surviving said travel (and for kicks only 240 of the known 500 amino acids are naturally occurring [1] and it depends entirely on random mutations and environment)? Self-replicating nanobots, one half of the holy grail (billions of years being the other) that would place in the hands of mortals that which allowed natural selection to give birth to the aforementioned aliens? At the technological level of relativistic weapon systems + nanobots you're basically talking the ability to do whatever you wanted. If they can have near omnipotent self-replicating nanobots flying to destroy civilizations, why not to slightly shift their stellar orbit by consuming the rest of the solar system (relativistic guidance and navigation is an entirely different ordeal) to add some interstellar non-determenacy or send out nanobots to replicate their civilization on hospitable solar systems? Sending out a bunch of relativistic nukes is probably going to be one hell of an expense (don't imagine you'd be able to do a gravitational sling shot without a black hole or perfect premonition). If the civilization was that scared of being wiped out, it'd probably be cheaper to build a single defense for their civilization rather than a bunch of attack platforms intended to wipe out other civilizations.

hnha 13y ago

or they could be the complete opposite, sad about our infighting and hate between our own. they might come to liberate us so we can live happier lives.

Houshalter 13y ago
That's very true. Although I think there is a good chance they could have incompatible values to us. It'd be pretty unlikely for two intelligent beings to evolve exactly the same morality. Imagine if they killed us all for eating meat, or if they were strong environmentalists like in The Day the Earth Stood Still. Or something else completely alien to us.