Alien neuropunk slaver civilizations

Here’s some blue-sky spec­u­la­tion about one way alien sapi­ents’ civ­i­liza­tions might de­velop differ­ently from our own. Alter­na­tively, you can con­sider it con­wor­ld­ing. Con­tent note: tor­ture, slav­ery.

Look­ing at hu­man his­tory, af­ter we de­vel­oped elec­tron­ics, we painstak­ingly con­structed ma­chines that can perform gen­eral com­pu­ta­tion, then built soft­ware which ap­prox­i­mates the work­ings of the hu­man brain. For in­stance, we nowa­days use in-silico re­in­force­ment learn­ing and neu­ral nets to solve var­i­ous “messy” prob­lems like com­puter vi­sion and robot move­ment. In the fu­ture, we might scan brains and then em­u­late them on com­put­ers. This all seems like a very cir­cuitous course of de­vel­op­ment—those al­gorithms have ex­isted all around us for thou­sands of years in the form of brains. Put­ting them on com­put­ers re­quires an ex­tra layer of tech­nol­ogy.

Sup­pose that some alien species’s biol­ogy is a lot more ro­bust than ours—their home­o­static sys­tems are less failure-prone than our own, due to some differ­ence in their en­vi­ron­ment or evolu­tion­ary his­tory. They don’t get brain-dam­aged just from hold­ing their breath for a cou­ple min­utes, and open wounds don’t eas­ily get in­fected.

Now sup­pose that af­ter they in­vent agri­cul­ture but be­fore they in­vent elec­tron­ics, they study biol­ogy and neu­rol­ogy. Com­bined with their ro­bust biol­ogy, this leads to a world where things that are elec­tronic in our world are in­stead con­trol­led by vat-grown brains. For in­stance, a car-build­ing robot could be con­structed by grow­ing a brain in a vat, hook­ing it up to some ac­tu­a­tors and sen­sors, then dos­ing it with happy chem­i­cals when it cor­rectly builds a car, and stim­u­lat­ing its no­ci­cep­tors when it makes mis­takes. This re­ward­ing and pun­ish­ing can be done by other lab-grown “over­seer” brains trained speci­fi­cally for the job, which are in turn man­u­ally re­warded at the end of the day by their owner for the to­tal num­ber of cars suc­cess­fully built. Cus­tom-trained brains could con­trol chem­i­cal plants, traf­fic lights, surveillance sys­tems, etc. The ac­tu­a­tors and sen­sors could be ei­ther biolog­i­cally-based (lab-grown eyes, mus­cles, etc., pow­ered with liquefied food) or pow­ered with com­bus­tion en­g­ines or steam en­g­ines or even wound springs.

Ob­vi­ously this is a pretty ter­rible world, be­cause many minds will live lives with very lit­tle mean­ing, never grasp­ing the big pic­ture, at the mercy of un­mer­ciful hu­man or vat-brain over­seers, with­out even the op­tion of suicide. Brains wouldn’t nec­es­sar­ily be de­signed or drugged to be happy over­all—maybe a brain in pain does its job bet­ter. I don’t think the own­ers would be very con­cerned about the eth­i­cal prob­lems—look at how hu­mans treat other an­i­mals.

You can see this tech­nol­ogy as a sort of slav­ery set up so that slaves are cheap and un­sym­pa­thetic and pow­er­less. They won’t run away, be­cause: they’ll want to perform their du­ties, for the drugs; many won’t be able to sur­vive with­out own­ers to top up their food drips; they could be de­vel­oped or drugged to en­sure docil­ity; you could pre­vent them from even get­ting the idea of eman­ci­pa­tion, by not giv­ing them the nec­es­sary sen­sors; per­haps you could even set things up so the over­seer brains can read the thoughts of their charges di­rectly, and pun­ish bad thoughts. This world has many par­allels to Han­son’s brain em­u­la­tion world.

Is this sce­nario at all likely? Would these civ­i­liza­tions de­velop biolog­i­cal su­per­in­tel­li­gent AGI, or would they only be able to cre­ate su­per­in­tel­li­gent AGI once they de­velop elec­tronic com­put­ing?