One of the most important points raised by the sequences is that not all minds are like humans. In quite a few places, people have discussed minds with slight changes from human minds, which seem altogether different. However, a lot of this discussion has been related to AI, as opposed to minds created by evolution. I'm trying to think of ways that minds which evolved, and are effective enough to start a civilization, could differ from humans'.

Three Worlds Collide would seem like an excellent starting point, but isn't actually very useful. As far as I recall, the Babyeaters might have learned their baby eating habits as a result of societal pressure. The main difference in their society seemed to be the assumption that people who disagreed with you were simply mistaken: this contrasts to humans' tendency to form rival groups, and assume everyone in the rival groups is evil. The Super-Happies had self modified, and so don't provide an example of an evolved mind.

So here are my ideas so far.

  • A species could have the same neural pathways for wanting and liking. This would lead to far less akrasia.
  • A species could have a different set of standards for boredom. This seems to be one of the most precarious values in the human mind.
What other ways can you think of?

New Comment
67 comments, sorted by Click to highlight new comments since:

Robert Freitas' Xenopsychology article has some interesting speculation around this theme. I recommend reading the whole thing, but here are some excerpts to get you interested:

Ganglionic intelligence:

On Earth, evolution favored voted the appearance of intelligence in two major classes of animal nervous systems, called ganglionic and chordate. It has its own peculiar psychology.

The invertebrates, representing perhaps 97% of all animal species alive today, took the ganglionic intelligence option. The earthworm is typical. Each of its many segments is almost an individual organism unto itself, having its own set of kidneys, muscles, sensors and so forth. Coordination is achieved by a thin latticework of nerve fibers crisscrossing from side to side and lengthwise. The ganglionic system resembles a ladder with bulbous neural tissues at the joints. Invertebrate organisms thus are comprised of a collection of sub-brains, each of which controls a separate part of the animal with fairly complete autonomy and no real centralized control. Sensors and their ganglia tend to cluster nearer the head, making not a true brain as we understand the term but rather a large bundle of distinct fibers. Such a nervous system is highly efficient for responding quickly to stimuli. Each clump of nerve cells becomes expert at some particular function–detecting and passing along sensory information, sweeping a leg or swing in wide uniform arc, opening and closing the jaws in slow munching motions during feeding, and so on.

Might extraterrestrials develop a high ganglionic intelligence that has, never developed on Earth despite hundreds of millions of years of opportunity? [...] It is hard for us to imagine the mentality of beings with advanced ganglionic intelligence. Dr. H. Chandler Elliot, a neurologist at the University of Nebraska College of Medicine, notes that humans normally disregard their internal organs. We respond to an empty stomach or a feeling of indigestion but normally we ignore its activities. Says Elliot: "The head of an insect apparently regards not only its viscera but also its legs, wings, and so on, with similar detachment. If one deftly clips off the abdomen of a feeding wasp, the head may go on sucking, obviously not distressed. The mind of such a creature, must be alien to us almost beyond comprehension." [...]

Yet another possibility is alien minds incorporating the advantages of both ganglionic and chordate architectures. For instance, each invertebrate sub-brain might evolve and enlarge to avoid multiplication of internal interconnections. This development is most likely in a radially symmetrical sea creature, wherein each brain has roughly equal access to sensory input and motor controls. Such creatures would have "collegial" mentalities, something akin to the many voting computers aboard the Space Shuttle, with multiple personalities within each organism and the ability to maintain consciousness under extreme physical trauma so long as any one brain remained functional.

Alien cultural universals:

To illustrate his point, he first refers to an inventory of the elements of human nature compiled by the American anthropologist George P. Murdock during a study of cultural universals:

Age-grading, athletic sports, bodily adornment, calendar, cleanliness training, community organizations, cooking, cooperative labor, cosmology, courtship, dancing, decorative art. divination, division of labor, dream interpretation, education, eschatology, ethics, ethnobotany, etiquette, faith healing. family feasting, firemaking, folklore, food taboos, funeral rites, games, gestures, gift giving, government, greetings, hairstyles, hospitality, housing, hygiene, incest taboos, inheritance rules, joking, kin groups, kinship nomenclature, language, law, luck superstitions, magic,. marriage, mealtimes, medicine, obstetrics, penal sanctions, personal names, population policy, postnatal care, pregnancy usages, property rights, propitiation of supernatural beings, puberty customs, religious rituals, residence rules, sexual restrictions, soul concepts, status differentiation, surgery, toolmaking, trade, visiting, weaving, and weather control.

Wilson insists that few if any of these elements are inevitable outcomes of either high intelligence or advanced social life; rather that "human nature is just one hodgepodge out of many conceivable." An entomologist by training, he has no trouble imagining an alien insectlike society whose members am even more intelligent and complexly organized than people, yet which lacks many of the qualities listed in Murdock's inventory. The alien inventory:

Age-grading, antennal rites, body licking, calendar, cannibalism, caste determinism, caste laws, colony-foundation rules, colony organization, cleanliness training, communal nurseries, cooperative labor, cosmology, courtship, division of labor, drone control, education, eschatology, ethics, etiquette, euthanasia, firemaking, food taboos, gift-giving, government, greetings, grooming rituals, hospitality, hosing, hygiene, incest taboos, language, larval care, law, medicine, metamorphosis rites, mutual regurgitation, nursing castes, nuptial flights, nutrient eggs, population policy, queen obeisance, residence rules, sex determination, solder castes, sisterhoods, status differentiation, sterile workers, surgery, symbiont care, toolmaking, trade, visiting, weather control . . . and still other activities so alien as to make mere description by our language difficult.

Alien emotions:

Of course, extraterrestrial sentients may possess physiological states corresponding to limbic-like emotions that have no direct analog in human experience. Alien species, having evolved under a different set of environmental constraints than we, also could have a different but equally adaptive emotional repertoire. For example, assume that human observers land on another and discover an intelligent animal with an acute sense of absolute humidity and absolute air pressure. For this creature, there may exist an emotional state responding to an unfavorable change in the weather. Physiologically, the emotion could be mediated by the ET equivalent of the human limbic system; it might arise following the secretion of certain strength-enhancing and libido-arousing hormones into the alien's bloodstream in response to the perceived change in weather. Immediately our creature begins to engage in a variety of learned and socially-approved behaviors, including furious burrowing and building, smearing tree sap over its pelt, several different territorial defense ceremonies, and vigorous polygamous copulations with nearby females, apparently (to humans) for no reason at all. Would our astronauts interpret this as madness? Or love? Lust? Fear? Anger? None of these is correct, of course the alien is feeling badweather.

... and more.

The ganglionic intelligence reminds me somewhat of Watts' Thing (link to his short story)

I'll give it a few shots...

Based off of something I once saw about octopuses, how about a creature with no sense continuity of self? Or no sense of self at all, even?

Or there's several ways to look at a creature that is largely immortal. I'm thinking of two specifically: One is so hearty that they can hardly suffer damage or death no matter what. The other is generally long-lived, but largely because they are careful (I think Eliezer talked about this somewhere, but I can't find it) that they would be terrified of crossing a street or driving a car because even a 1 in 100 million chance of death is too dangerous (but they would expect to get a royal flush several times over the course of their lives, they live that long).

I like the idea for wanting/liking being the same thing. You could play off of that theme with stuff like:

  • creatures who have conscious control over their reward circuitry
  • creatures whose neural reward pathways grow with age, so they experience rewards more deeply the older they get
  • creatures whose pathways for liking are transmitted to others, so the more creatures that like something, the deeper the pleasure
  • creatures whose desires can be traded

That's about all I could come up with in the past ~20 minutes. Hope they help.

The nearly immortal and very careful species is a trope that shows up in some science fiction stories. See for example the Puppeteers in Niven's Ringworld. Note also that there's some empirical evidence for this on a small scale. Countries which have longer lifespans generally have more risk averse populations. There's an argument that dueling went out of style in many cultures about the same time that the likelyhood to die from a lot of diseases went down as well. I don't know if the data really reflects that claim though.

[-]Jack100

There are the aliens in Blindsight. Rot13 Spoiler: Jub ner rkgerzryl vagryyvtrag ohg ynpx pbafpvbhfarff.

Speaking of Blindsight, Watts drew heavily on Metzinger's Being No One. One of my favorite Watts blog posts was his post on the PRISM model of consciousness http://www.rifters.com/crawl/?p=791 and to quote from the PRISM guy:

“…One could imagine a conscious nervous system that operates as humans do but does not suffer any internal strife. In such a system, knowledge guiding skeletomotor action would be isomorphic to, and never at odds with, the nature of the phenomenal state — running across the hot desert sand in order to reach water would actually feel good, because performing the action is deemed adaptive. Why our nervous system does not operate with such harmony is perhaps a question that only evolutionary biology can answer. Certainly one can imagine such integration occurring without anything like phenomenal states, but from the present standpoint, this reflects more one’s powers of imagination than what has occurred in the course of evolutionary history.”

I'd guess because pain has to be immediate to be of value, so the more processing you heap on it the less useful it becomes; and species tend to evolve pain before they evolve utility-judging systems.

Awesome link!

I'm not quite sure if I think that the PRISM theory is particularly accurate though. There are plenty of animals that can mediate between conflicting impulses, which don't seem very conscious. Like cows, or frogs. It seems plausible that people may have already built robots that do this.

I'm fairly capable of imagining living in my body while everything below my neck exhibits goal-oriented behavior without my input. I think I would still be conscious in that state.

So I don't really think that consciousness is the phenomena of mediating between conflicted motor impulses.

There are plenty of animals that can mediate between conflicting impulses, which don't seem very conscious. Like cows, or frogs.

What makes you think that? I'm not entirely sure of frogs, but cows seem obviously conscious to me.

I'm with Kaj on this. One man's modus ponens is another man's modus tollens. I don't mind counting frogs or cows as partially conscious to the extent they reconcile conflicting impulses, and I don't understand why you take such a viewpoint as a refutation of PRISM. Why is consciousness all-or-nothing, rather than a matter of degrees?

It's definitely a matter of degree, but I don't think that humans being more conscious is a result of being more able to reconcile conflicting impulses. I think of the refutation as more being in my second paragraph.

Imagine if a friend of yours suddenly somehow lost all control of their body except for their mouth. They do things, and have introspective access to why, and can talk about what they're doing, but take no part in actually deciding what to do.

I would still consider them to be conscious, and intuitively think that something that is more able to talk about itself is more conscious, more so than I think that being able to decide makes something conscious.

Human consciousness can mediate internal conflicts, but I don't think that internal conflict mediation is a sufficient condition for consciousness.

Imagine if a friend of yours suddenly somehow lost all control of their body except for their mouth. They do things, and have introspective access to why, and can talk about what they're doing, but take no part in actually deciding what to do. I would still consider them to be conscious, and intuitively think that something that is more able to talk about itself is more conscious, more so than I think that being able to decide makes something conscious.

The degree vs kind distinction still works for that. You're just proposing a slippery slope and denying the first step is downward. Keep taking your friend and chopping away capabilities (more capabilities means more possibilities means more potential for conflict) - chop away their long-term memory, short-term memory, random senses, etc.

Where do you say they begin to lose some consciousness? I'm happy to say that they lose a little bit every time, with some losses being reparable over time and the exact loss.

(And I wonder how conscious your friend would actually be. Certainly, there's a lot of potential there, but judging from float tanks and the psychological effects of isolation tanks...)

(Notes confusion)

Intuitively, I think they stop being conscious when they stop being able to talk to me about what they subjectively think about the world.

Not being able to control your body makes you a bit less conscious, but not nearly as much as removing long term memory. I don't think that the degree of conflict is as important as the degree of representation.

Confusion on "subjectively think":

I think that this is a proxy for having an experience of the world, qualia and that sort of thing.

Confusion on being "able to talk":

I can have an inner dialogue without opening my mouth and vocalizing to other people, and I still report consciousness. If I had magical telepathy powers that let me access someone's inner dialogue without them talking, then they would probably be able to convince me of their consciousness.

Those are really just the ideas that I'm using to think about this. Since they seem really important to what I personally mean when I say consciousness, I don't think its the ability to mediate internal conflicts.

(And I wonder how conscious your friend would actually be. Certainly, there's a lot of potential there, but judging from float tanks and the psychological effects of isolation tanks...)

I feel like the losing consciousness is probably a result of losing sensory input on which to base a model of the world. When your world model is gone, its harder to talk about things.

You only have as many qualia as you need to; sensory data is discarded as much as possible. (Look at meditation, how much one experiences but does not notice. Look at dreams - they seem vivid and real, until one tries to see specific detail like reading written material.) And what one perceives is strongly shaped by what one expects (eg. the ba-ga experiment or the entire prediction-is-intelligence line of thought - On Intelligence comes to mind). Look at how the mind shuts down when there is little to do, in things like highway hypnosis.

(Maybe you should read the PRISM papers.)

Upon reading the papers, it seems like were talking about different things.

I was talking about what I thought consciousness was (like, what I would label as conscious and unconscious), and I think you were talking about what it does/is for.

Is there a difference between what something is and what something does?

I've got to agree, Blindsight is an excellent example. It successfully (IMHO) portrays aliens that are clearly quite intelligent, and yet utterly inhuman.

SPOILERS SPOILERS

The most frightening thing about them is that none of the usual tropes apply to them; mutual understanding and/or empathy isn't an option because their psychology (or possibly lack thereof) makes it completely impossible. There's nothing there to empathize with.

[-][anonymous]00

I think you meant Rot13.

[-]Jack00

Yes. The two keys are right next to each other on my keyboard :-)

Look into eusocial animals, e.g. ants, bees, wasps, termites, and naked mole rats. Intelligent minds of similar evolutionary origin would be very different from ours.

That depends on whether they're intelligent on an individual or a hive level. After all, we are in a sense eusocial collections of cells.

An interesting exercise would be to try to imagine a society of humans whose minds operated the way most intellectuals today imagine human minds operate.

For instance, see the recent post by LukeProg on values being relative. What would people be like if they really did have states as terminal values, the way folk psychology says they do?

Another exercise is: What would society be like if everyone knew how their minds operated? How would we act if people didn't believe in free will? (Asking that question gives me a headache... it seems to imply that we have free will.) How would we act if we were rational utility-maximizers?

Tweak the bandwidth of information exchange and see what you get. With weaker communication powers, you get most nonhuman social animals on Earth, from ants to wolves to prairie dogs: able to exchange meaningful information ("food here"; "predator there") but not to express abstractions. How about with higher-bandwidth or more powerful information exchange? Suppose that I could, in the space of a few seconds of speech, convey not merely a fact, but the equivalent of paragraphs of information as to how I knew that fact?

(Or, of course, I could lie. Is increased bandwidth really all that useful? If so, why don't we have it?)

Alternately:

Some have suggested that a human mind is a sort of symbiote between genetic replicators (the human body including brain) and memetic replicators (culture, language, social behaviors, ideas). Certain aspects of human behavior are relatively constant across cultures, while others are highly variable. For instance, there are many different languages, but they all fit certain common patterns, because they have to be learned by a human brain.

Suppose that the relative strength of cultural-memetic replicators is increased: a species in which any given biological individual can be expected to participate not only in several different cultural milieux over the course of its life, but where a process somewhere between religious conversion and career-change takes the significance that reproduction has for humans. Converting other adults to follow your religion/worldview/profession/habits/hobbies/quirks strongly dominates reproduction and child-care as demands on organisms' energy. Rather than a reproductively successful individual being one who has several children, and a "loser" being someone who cannot find a mate, a successful person of this species is expected to have several followers or students in different aspects of life, and a "loser" is someone who is not imitated by anyone in any regard.

Spread of your genes is what's optimized for by evolution - all of our values are conditional on the reality of how we live and spread our genes. Hive creature? Different values. Solitary creature? Different values. Underwater? Values. Cannibalistic? Eating babies is an important value, it's unwise to make the relativistic claim that it's just "social" - that's a slippery slope to societies with no babyeating at all.

Relative rate of thinking. The universe may appear to be very different to very fast or slow thinkers relative to humans.

Also, relative rate of return on investment, and relative lifespan.

Ooh, I love this kind of thing! (i also consider myself very good at it) However, I can't really communicate somehting as complex as the kind of idea that makes a non gimmicky, truly alien mind through static text, mainly because I suck at explaining things that way. And I am confident the results will be much better if it us customized to the story anyway. So, I suggest we meet up over IM for a brainstorm session. As I said, I love this kind of stuff and will do it just for fun without you needing to mention me in the credits or anything like that if you don't want to.

[-]djcb30

Ah, that's a very intruiging question - to which we probably won't get any better answer that what we can come up with through speculation...

It's really, really hard to extrapolate from n=1 intelligent species, but as you suggest, there may be some universals; some things I would consider in coming up with a credible alien intelligence:

  • map intelligence vs mind properties for various earth species, and consider how the properties could change if you'd have a much more intelligent species (i.e., a more intelligent species may have more introspective control over things like boredom, akrasia and so on.)
  • consider how the natural enviroment may influence the group size, amount of coordination needed and so on

Then there's the fact that certain traits may make it much more likely to encounter aliens that have them - if they don't have the need or curiosity to go beyond their home ground, it's much less likely you would encounter them.

I feel however that it is really hard to think about so very different minds. While I don't subscribe to his non-reductionist views of the mind, Thomas Nagel's What is it like to be a bat? is a classic in this area, as it's reprinted in The Mind's I.

Anyway, good luck writing your fiction!

You have read A Fire Upon The Deep, right?

Alternately, imagine being able to explicitly fork and merge streams of cognition as easily and explicitly as one can fork code in git. You could imagine each stream gets its own set of permissions, so the left hand might really not know what the right tentacle is doing.

Humans (roughly) desire to be happy. Imagine aliens who desire to be angry instead.

Humans (roughly) desire to be happy. Imagine aliens who desire to be angry instead.

Isn't it the act of desiring that makes it happiness, not the name?

Only sort of.

While we are happy we have certain mental dispositions which are unrelated to the ones we desire when we desire happiness. Happy people are more likely to give to charity. Happy people (or at least people on certain antidepressants) are also more likely to judge harming others as wrong -- whether this translates into less actual harm, I don't know. Further, people tend to prefer that others be happy -- that's why people are always telling depressed people to smile.

Imagine that in these aliens, happiness worked in terms of its behavioral effects as it does in humans, as did anger. But aliens might still want themselves to be angry even as they want others to be happy.

Further, people tend to prefer that others be happy -- that's why people are always telling depressed people to smile.

Yes, but also smiling makes people happier. So it is pretty good advice for depressed people.

But aliens might still want themselves to be angry even as they want others to be happy.

By this you mean that when these aliens get what they want, instead of feeling content and magnanimous, they feel discontent and destructive? Don't you see how that could be an evolutionary handicap?

I'm not sure anger necessarily leads to discontent. If you want to get mad as hell and break things, you might feel better after doing so if you've been stifling a lot of discontent up to then. I don't really pretend to know what's going on in the head of a rioter but this seems possible.

I don't mean that actions anger predisposes you toward leave you discontented, I mean that part of the physiological emotion of anger is discontentment.

Yeah, I guess it could be a handicap. But then human brains have so many bizarre flaws and spandrels that I think we could get away with swapping a few of ours for this one.

In this post Robin Hanson presents a plausible vaguely anthropic argument that alien psychology (as well as physiology) may be more like human then you'd expect.

It's hard to imagine something stranger than things that already exist, like eusocial insects, slime molds, solitary mammals, total winner-take-all reproduction, parasites, or species that don't care for or recognize their young.

[-]Jack20

A species that evolved intelligence but for which the social brain hypothesis is false might be very different-- but I don't know if it is plausible that such a species would develop a civilization.

Do you mean what Eliezer calls the Machiavellian intelligence hypothesis? (That is, human intelligence evolved via runaway sexual selection--people who were politically competent were more reproductively successful, and as people got better and better at the game of politics, so too did the game get harder and harder, hence the feedback loop.)

Perhaps a species could evolve intelligence without such a mechanism, if something about their environment is just dramatically more complex in a peculiar way compared with ours, so that intelligence was worthwhile just for non-social purposes. The species' ancestors may have been large predators on top of the food chain, where members are typically solitary and territorial and hunt over a large region of land, with its ecosystem strangely different from ours in some way that I'm not specifying (but you'd need a pretty clever idea about it to make this whole thing work).

These aliens wouldn't be inherently social the way humans are, but they wouldn't be antisocial either--they would have noticed that cooperation allows them to solve even more difficult problems and get even more from their environment. (Still in a pre-technological stage. Remember, something about this environment is such that it provides a nearly smooth, practically unbounded (in difficulty) array of puzzles/problems to solve with increasing rewards.) Eventually, they may build a civilization just to more efficiently organize to obtain these benefits, which will also allow them to advance technologically. (I'm probably drawing too sharp of a distinction between technological advancement and interaction with their strangely complex environment.)

They might lack the following trait that is very central to human nature: affection through familiarity. When we spend a lot of time with a person/thing/activity/idea, we grow fond of it. They might not have this, or they might not have it in such generality (e.g. they might still have it for, say, mates, if they reproduce sexually). They might also be a lot less biased than we are by social considerations, for the obvious reason, but perhaps they have less raw cognitive horsepower (their environment being no substitute for the human pastimes of politics and sex).

Recklessly speculative, obviously, but I gather that's all we can hope to offer to Solvent.

How about:

  • Enhancing the brain with hardware for that memory/intelligence boost.
  • Superior intuition for probability
  • Different/worse optical illusions/blind spots from different optical nerve wiring
  • Superior 3D spatial visualization for ocean dwellers
  • Creatures that get stuck in a correlation/causation confusion and manage to continue along mostly successfully
  • Types of synaesthesia far more elaborate than in Humans
  • A conflict between intelligence and instinct where instinct is in total control but intelligence knows it's doing the wrong thing
  • Minds partitioned more finely than the human left/right hemisphere split, with more independence between them
  • Species that specialize their intelligence like insects becoming workers/warriors/queens/drones etc. but for intelligence

Lots of fun possibilities in this area.

Species that specialize their intelligence like insects becoming workers/warriors/queens/drones etc. but for intelligence.

Humans do this (and increasingly so). e.g. we have physicists, mathematicians, economists, psychologists etc.

[-][anonymous]00

Perhaps one might imagine a creature that essentially cultivates multiple personalities disorder for each task they perform, where the task at hand is (more or less) their utility function, unless something like survival instict kicked in an override?

Imagine a creature that didn't do that!

An alien species might be very adaptive, able to learn new things exclusively - and efficiently - by media such as audio, video, and text. Such a species might not have its culture organized around the seizing of others and agglomerating their brains and the defense from likewise. Prictualcre to determine who eats from whom, and how much brain to take, and how much work must be put into studying directly from books, and concomitant social enforcement mechanisms regulating such might be entirely absent. At first glance, that seems like a lot, though not everything that makes us who we are, but a bit of thought will reveal how deeply prictualcre is embedded in not just our culture but our biology, and in fact a species without it would be almost unimaginably strange. We simply don't notice it because of its universality on biological, cultural, and other levels.

My impression is that convergent evolution is commonplace. Boredom or equivalent algorithms are necessary for any complex learning. Emotions are molded by game theory. Once you have an intelligent social animal the selection pressures relentlessly shape the mind. If I met an alien I wouldn't be surprised if he or she was almost exactly like me in a lot of ways, to the extent that he or she would describe similar drug experiences and have similar archetypes in his or her fiction. In their vast diversity humans vary about as much as biological minds can vary. Think of it as a challenge: I'd be very interested in alien minds that clearly took these considerations into account but still ended up truly bizarre.

On the other hand, the aliens could (for example) have emotions encoding a different game-theoretic strategy, or a different prior over what games it's going to have to play.

For example, boredom is useful in a hunter gatherer, but in an ambush predator it would be counter-productive as their optimum strategy might be remaining in the same place for long periods.

Do felines not experience boredom?

Interesting question. They enjoy playing, but they also enjoy lying in one place for hours at a time.

Is boredom with an activity different from not enjoying the activity? Are zombies bored? I would say they are not.

Zombies are fake.

[-]Jack30

In their vast diversity humans vary about as much as biological minds can vary.

Can you be more precise about your reasons? I'm sure they're in the linking articles somewhere but there's a bit too much there for me to comprehend.

Sorry, I didn't mean to imply that there were already existent strong arguments for that; it's just my personal speculation. I think it's one of those things where I could type for a few days straight and still not accurately describe my intuitions and their origins. Better to just read a ton of vaguely relevant stuff, I think.

I wonder what, if any, the relationship is between several things: predicting what humans will learn about the laws of nature, and predicting what alien life is likely to be like, for both an estimate and confidence in it.

It feels incongruous that I am relatively agnostic and uncertain about alien life's form (assuming it exists) but somewhat confident physics is close to hitting bedrock within the century.

I could just be ill-informed and reasoning poorly. Even so, I would expect patterns to those shortcomings.

Honestly, the odds of a planet that produces social talking thinkers not producing ones deeply similar to humans are pretty small.

[-][anonymous]30

Why do you say that?

Because natural selection leads to convergence. Yes, mindspace is (probably) infinite- but the space of minds reachable by natural selection is a pretty small subset. I can confidently predict that creating progeny will be important to them, that they'll form pair bonds, that there will be sexual jealousy, that the sex with larger gametes will have a history of being seen as secondary, that status will motivate them, and so on. Solitary intelligences are astoundingly unlikely compared to social intelligences, and I'm counting hiveminds as solitary rather than social. (A queen bee just needs to stab its sisters to reproduce, not seduce the highest-status male.)

Speculative fiction gets a lot less fun when you enforce practicality tests. The Left Hand of Darkness comes to mind as a marvelous piece of fiction, but lampshades the existence of ambisexuals, by supposing they were an experiment left on the world by an unknown entity, since it wouldn't make evolutionary sense for them to exist.

Because natural selection leads to convergence. Yes, mindspace is (probably) infinite- but the space of minds reachable by natural selection is a pretty small subset. I can confidently predict that creating progeny will be important to them, that they'll form pair bonds, that there will be sexual jealousy, that the sex with larger gametes will have a history of being seen as secondary, that status will motivate them, and so on.

Ok. But there are a lot of problems with this. Consider for example a hermaphroditic species. Yes, they exist in nature. Or consider a species that reproduces asexually while shuffling their own DNA. There's at least one not too stupid species which does that on our planet. The gamete claim is hard to define carefully, but one should note that many primitive societies of humans have fewer not more gender distinctions. And there are matriarchal societies among humans. See for example the Nair. Among some vertebrate species this isn't just something that occurs in some groups but occurs universally in some species. See for example the spotted hyena, although I should warn you that reading about some of the details of their reproductive behavior can be potentially disturbing.

I've got a lot more examples where this comes from, like humans forming group marriages and other fun stuff. But there's a more fundamental problem here. You are looking at some of only the more obvious potential universals. Even those aren't universals. But there are so many things that humans have that are universals that could easily change and could have drastic impacts on societies. Kaj's post above includes among other things a nice excerpt of a very long list of human universals that one wouldn't even likely recognize as such. Changing even a small number could have drastic consequences. Indeed, most of that list is so ingrained in humans that when we write about alien species we often include most of them. Granted, most alien species are not constructed with much thought put into it (the rubber-forehead problem) but the pattern should be striking.

Yes, they exist in nature.

To the best of my knowledge, no simultaneously hermaphroditic species are social (I'll use pack formation as the threshold for social). Do you think that's a coincidence?

You are looking at some of only the more obvious potential universals.

I didn't want to write a page; as you point out, Kaj had already quoted one.

Changing even a small number could have drastic consequences.

My argument is that those consequences are likely to be drastically negative. Remember, most mutations are deleterious!

In order to have a society where X is different, you need to identify a new evolutionarily stable equilibrium. To use The Left Hand of Darkness again as an example, there's a very low speed limit enforced by psychology- no one sees a reason to go faster. That's just unbelievable- there are no crazy people who don't have trouble driving faster? Those people don't get wealthier because they make trips faster / have a longer driving season (the planet, Winter, has terribly cold winters which shut down all travel, and so being able to drive faster means being able to leave later), causing them to reproduce more than more timid drivers?

And even if an alien species manages to make it to human-level intelligence without a calendar (perhaps their planet doesn't have a satellite or axial tilt), will that invalidate the claim that they'll be deeply similar to humans? Hardly. They'll still have the imprint of their machiavellian evolution all over them.

To the best of my knowledge, no hermaphroditic species are social (I'll use pack formation as the threshold for social). Do you think that's a coincidence?

I think that's a function of a variety of issues. One is evolutionary lock-in. Some things are more difficult to alter than others. Changing your basic reproductive system is tough because so much can go wrong. There are very few parthenogenetic species. But they do seem to show up in clumps. There are a lot of lizards in that category and lizards generally don't form packs. But if mammals had ended up with the necessary genetics to be easily engage in parthenogenesis things could easily have looked very different, and I see no reason why versions of us wouldn't be having a very similar conversation, with one of us insisting that the flexibility given by parthenogenesis as an option is necessary to get sufficient intelligence.

I'm also curious if you think this sort of logic applies also to your claim about gametes, given that the example I gave was an unambiguously social species. Do you agree that the gamete claim is wrong?

Changing even a small number could have drastic consequences.

My argument is that those consequences are likely to be drastically negative. Remember, most mutations are deleterious!

Most mutations are probably neutral. It is probably true that most mutations which have some selection probably have a negative selection pressure. However, this remark confuses why mutations are so often negative. Most species have not only adopted to specific niches, their genes function in an intertwined mesh. So if I mutate one gene, all the genes that interact with that gene are potentially unhappy. That's not a problem when one is discussing species evolving wholesale. They are then free to include or not include what we see as universals.

Those people don't get wealthier because they make trips faster / have a longer driving season (the planet, Winter, has terribly cold winters which shut down all travel, and so being able to drive faster means being able to leave later), causing them to reproduce more than more timid drivers?

Huh? I'm not sure I follow this example. You mean driving a car? Humans have barely had a hundred years of car travel, way too little to show selection pressure. Moreover, driving is dangerous so there's an obvious tradeoff. And many people don't drive regularly at all (I for example live in a city with a decent public transit system). We also have laws which are enforced against extreme driving. And as long as there is other traffic you won't gain too much from being able to drive faster since you'll be in the traffic jams. Moreover, people do drive at drastically different speeds. Finally, and most damaging, cultural norms about how fast to drive have changed albeit slowly, and they vary a lot from country to country. Some places have speed limits as high as 160 km/hour (99 miles an hour). Around 1900, 25 miles an hour was considered to be blindingly fast for a car. It was a big deal in 1896 when parts of England has the speed limit raised to 14 miles an hour. If you had a species that had better reflexes and hand-eye coordination than humans have I could easily see 200 mph as the speed limit. Similarly, if human coordination was poorer or humans were more risk averse I could easily see us keeping the early 20th century speed limits.

Remember, just because we've hit some set of evolutionarily stable equilibria doesn't mean that there aren't a lot of others out there. We know there are a lot out there because there are a lot of other species. We don't know how many there are that support highly intelligent life, but the existence of parrots, dolphins, elephants and some of the brighter corvids strongly suggest that there's a lot of room.

And even if an alien species manages to make it to human-level intelligence without a calendar (perhaps their planet doesn't have a satellite or axial tilt), will that invalidate the claim that they'll be deeply similar to humans? Hardly. They'll still have the imprint of their machiavellian evolution all over them

Are you trying to argue that it is likely that intelligent, social species will have a lot of ability to trick each other and engage in clever schemes and have sophisticated theories of mind? If so, I agree that seems very likely. But none of the universals you gave are functions of that feature.

I see no reason why versions of us wouldn't be having a very similar conversation, with one of us insisting that the flexibility given by parthenogenesis as an option is necessary to get sufficient intelligence.

The primary reason I see is the Machiavellian Intelligence hypothesis- if human-level intelligence is reproductively successful primarily to seduce and outwit, then a species that does not need to seduce or outwit in order to get the best partners will not develop human-level intelligence.

My point in this debate is I want to see the math. The first proposed version of group selection sounds plausible but isn't once you do the math.

I'm also curious if you think this sort of logic applies also to your claim about gametes, given that the example I gave was an unambiguously social species. Do you agree that the gamete claim is wrong?

I am not a biologist, and so am taking it on the advice of experts that gamete size determines social role in a bisexual society. I believe the strength of that role depends on the relative size of the gametes- and so it may be that hyena gametes are very close in size or there is some other reason why they are an exception. I do agree that "secondary" was not the right way to put my claim- instead, I'll reword it to be that "in a bisexual society, sex roles will have deep similarity to human male-female roles."

That's not a problem when one is discussing species evolving wholesale. They are then free to include or not include what we see as universals.

Indeed, we could imagine an alien species that, for whatever reason, doesn't trade with one another. They're easy to imagine because they evolved on Earth. What I find entirely implausible is that such a species could be the first to make it to human-scale intelligence and technological domination.

We don't know how many there are that support highly intelligent life, but the existence of parrots, dolphins, elephants and some of the brighter corvids strongly suggest that there's a lot of room.

Right, but the differences between corvids and human-precursor primates are mostly superficial. If corvids had made it to human-level intelligence first, the similarities would be very deep.

The primary reason I see is the Machiavellian Intelligence hypothesis- if human-level intelligence is reproductively successful primarily to seduce and outwit, then a species that does not need to seduce or outwit in order to get the best partners will not develop human-level intelligence.

A hermaphrodic species could still need intelligence to outwit other members of the species.

My point in this debate is I want to see the math. The first proposed version of group selection sounds plausible but isn't once you do the math.

Huh? You made no mention of math earlier. Is there math you think you have to support your position? If so, yes, you are correct that there are game theoretic models that predict certain classes of behavior being more likely. For example, one does in fact expect certain patterns based on which gender invests more resources in the young. There are some really interesting examples where the males have for various reasons come to invest more in the young, and exactly what you expect often occurs, the females end up having harems of males and try to pump out as many offspring as possible. Jacanas are a good example of this

I believe the strength of that role depends on the relative size of the gametes- and so it may be that hyena gametes are very close in size or there is some other reason why they are an exception. I do agree that "secondary" was not the right way to put my claim- instead, I'll reword it to be that "in a bisexual society, sex roles will have deep similarity to human male-female roles."

The problem you are running into is that gamete size is only a very rough proxy for level of resource investment. And as a proxy it becomes weaker the more time the species spends raising its young. Your second comment is more accurate.

But it is important to realize that for all of this humans actually show major exceptions to the rules that one expects to have. In particular, one expects males to spend more resources engaging in showy demonstrations of fitness. And in fact in many species one does see this (peacocks are of course the most famous example). But in most human societies, females spend as much or more resources than males looking nice. Jewelry is generally much more common among human females than human males. This is strongly not what one would expect.

Indeed, we could imagine an alien species that, for whatever reason, doesn't trade with one another. They're easy to imagine because they evolved on Earth. What I find entirely implausible is that such a species could be the first to make it to human-scale intelligence and technological domination.

I agree with you that trade is probably a necessary precursor. I'm not sure that it is absolutely necessary for intelligence but it is presumably necessary for the technological domination part simply because no one is going to be able to develop all the technologies on their own. And in fact, trade behavior occurs in a variety of species in limited contexts.

Right, but the differences between corvids and human-precursor primates are mostly superficial. If corvids had made it to human-level intelligence first, the similarities would be very deep

Really? Some birds are able to sleep with only one side of their brain sleeping at a time, while the other half watches for predators. I don't know of any corvids that can do that, but it isn't at all implausible. You don't think that ability wouldn't drastically alter perception of self, and many other behavioral attitudes? And when one realizes that corvids lay eggs rather than giving live birth, how many issues related to that go away? Or if we turn to the list that Kaj helpfully supplied and look at dolphins, one of the other very intelligent species I listed, how many of the things on that list would simply not be important to them just because they live underwater? Firestarting and fire rituals would be out for obvious reasons. And that's just the most obvious starters. And then there's all the stuff on the list that is simply a product of humans being overactive pattern seekers (e.g. luck, magic, faith healing...) or having poor introspection ability (soul-concept), etc.

A hermaphrodic species could still need intelligence to outwit other members of the species.

It could. But do you agree that parthenogenesis is a very unlikely reproductive method for a HLI? (I've gotten tired of writing human-level intelligence.)

You made no mention of math earlier. Is there math you think you have to support your position? If so, yes, you are correct that there are game theoretic models that predict certain classes of behavior being more likely.

I thought it was obvious, and have paid for that mistake with karma.

In particular, one expects males to spend more resources engaging in showy demonstrations of fitness. And in fact in many species one does see this (peacocks are of course the most famous example). But in most human societies, females spend as much or more resources than males looking nice.

Let's take a step back here. Do you really believe that women spend more time and effort demonstrating their fitness than men do? Or are you trying to prove me wrong?

I think it obvious that men compete more than women do, that their competitions are riskier, and that human displays of fitness are status-based, and thus only partially visual (and then typically have more to do with posture than ornamentation). Indeed, the primary explanation for peacock tails appears to be the handicap they impose, not that they're visually stunning (although some argue that the eyes on the tails might hypnotize females). I'm confused why I even have to point that out. I agree that women also compete for men, and one of the primary ways they do that is looking nice. But male and female psychology line up with what you would expect from females investing more into children than males, and differences in family structures between regions match what you would expect from the environment.

Jewelry is generally much more common among human females than human males. This is strongly not what one would expect.

Men are more visual, because the health of a mother is important and primarily communicated through appearance; the primary long-term value of a man to a woman is his ability to provide material and social comfort. Thus, one would expect men to provide women with expensive and pretty tokens of affection.

You don't think that ability wouldn't drastically alter perception of self, and many other behavioral attitudes?

It might, but I don't know how much it would. I would look at split-brain patients and extrapolate from there- so there would probably be some oddness, but nothing fundamentally different.

And when one realizes that corvids lay eggs rather than giving live birth, how many issues related to that go away?

What issues related to live birth are you talking about? (For example, development time in womb compared to out of womb appears to be related to the limits that vaginal size puts on head size- I don't know if that favors live birth or eggs, but might be a limiting factor for eggs.)

how many of the things on that list would simply not be important to them just because they live underwater?

Is an underwater species likely to reach HLI before an abovewater species? My understanding is the metabolic demands of intelligence are high, and that appears to favor abovewater species. (Note that our planet, at 70-30 water-land, would seem to favor the species that had access to more of the surface area, suggesting that something else puts land ahead of water.)

And then there's all the stuff on the list that is simply a product of humans being overactive pattern seekers (e.g. luck, magic, faith healing...) or having poor introspection ability (soul-concept), etc.

But, how can you have an intelligence without pattern-seeking? Why would good introspection ability survive Machiavellian evolution?

It could. But do you agree that parthenogenesis is a very unlikely reproductive method for a HLI?

How unlikely is very unlikely? I'd agree that our evidence does suggest that I'd expect it to be less common than conventionally reproducing species. But that's not the point. Parthenogenesis is but one example of many different features which show up in fairly smart species on Earth. I only need to change a few to drastically alter what one would expect.

Let's take a step back here. Do you really believe that women spend more time and effort demonstrating their fitness than men do? Or are you trying to prove me wrong?

I don't know if time is the best metric in this sort of context, and I really don't know which gender generally spends more time.

In any event you seem to be missing my point so I will be a bit more explicit: In the vast majority of species that have substantial mate competition, the competition is almost exclusively among the males and it frequently takes a visual component. In humans that's false. We're an exception in that regard.

It might, but I don't know how much it would. I would look at split-brain patients and extrapolate from there- so there would probably be some oddness, but nothing fundamentally different.

Split-brain patients are a tiny fraction of the general population. The situation here also isn't that similar because it is a situation where they can turn off either half, or both, and the two communicate.

What issues related to live birth are you talking about?

Human universals surround pregnancy and the related issues. Almost every society has associated rituals and taboos. And some gender issues come from the fact that females are stuck for months being very vulnerable.

Is an underwater species likely to reach HLI before an abovewater species? My understanding is the metabolic demands of intelligence are high, and that appears to favor abovewater species.

I don't think that the metabolic issues will matter much. The total caloric intake of a dolphin or other large sea creature is quite large. The difficulty with making useful underwater tools would strike me as a more substantial problem. Dolphins do make some makeshift tools (such as some pods using sponges in front of their faces to protect themselves from spines when they go after spiny undersea life) but it seems substantially more difficult to actually go and make tools (for example, chipping stone would nearly impossible).

But, how can you have an intelligence without pattern-seeking?

Pattern seeking is one thing. Being an overactive pattern seeker is another. I would guess for example that if HLI had arose from elephants rather than primates one would see much less overactive pattern seeking because of the lack of threats they face.

Why would good introspection ability survive Machiavellian evolution?

Why wouldn't it? If I understand how minds work better that makes me more adept not less adept at manipulating others. There are arguments that this would not be the case, but they seem contingent on specifics of how humans function.

How unlikely is very unlikely?

I am poorly calibrated at assigning numbers to probabilities like this, so I don't think I can add more here.

I only need to change a few to drastically alter what one would expect.

My original claim was not that any noteworthy changes were impossible, but that any aliens that are social talking thinkers will be deeply similar to humans with high probability. (Implicitly attached is the claim that anything that isn't a social talking thinker we wouldn't consider intelligent and/or wouldn't reach HLI.)

It could be that some aspect of human psychology that was adaptive for us just fails to exist in some alien species. But I think the chances of that are amazingly small. It could be that something I think is an universal driven by practicality turns out to be a close competitor with something else (like sequential hermaphroditism vs. bisex, or whatever that's called), and on some planets things went the other way. If that's the case, I suspect deep similarities will still exist, and the change will be relatively minor. (For example, I suspect sequential hermaphroditism would be female->male, and the resulting social system would look a lot like pederasty.)

In the vast majority of species that have substantial mate competition, the competition is almost exclusively among the males and it frequently takes a visual component. In humans that's false. We're an exception in that regard.

Females don't compete in other species?

Human universals surround pregnancy and the related issues. Almost every society has associated rituals and taboos. And some gender issues come from the fact that females are stuck for months being very vulnerable.

So, there will be egg rituals and taboos instead, and I imagine some couples will trade off being 'pregnant'. The invention of incubators will be comparable to the (future) invention of artificial wombs.

I would guess for example that if HLI had arose from elephants rather than primates one would see much less overactive pattern seeking because of the lack of threats they face.

Alright, but that's an argument that elephants are less likely to reach HLI, as there's less selection pressure for pattern-matching / swiftness of thought.

Why wouldn't it? If I understand how minds work better that makes me more adept not less adept at manipulating others. There are arguments that this would not be the case, but they seem contingent on specifics of how humans function.

There are three factors in play: ability to know your motives, ability to lie convincingly, and ability to detect sincerity. The easiest one to drop is the ability to know your motives- when people promise to always be faithful, for example, they typically mean it, even though, beneath their consciousness, they don't.

Females don't compete in other species?

Note I made a comment about the vast majority of species. There's a conference devoted to such competition precisely because they are the exception not the rule.

Alright, but that's an argument that elephants are less likely to reach HLI, as there's less selection pressure for pattern-matching / swiftness of thought.

Speed of thought isn't necessarily related to overall intelligence. Note that elephants are one of the smarter species even though they aren't subject to much of the specific selection pressure that makes humans overactive pattern seekers.

There are three factors in play: ability to know your motives, ability to lie convincingly, and ability to detect sincerity. The easiest one to drop is the ability to know your motives- when people promise to always be faithful, for example, they typically mean it, even though, beneath their consciousness, they don't.

I don't see why knowing your motives is easier to drop than the ability to lie convincingly. That may be the general solution for most humans but that doesn't mean it is the easiest. Indeed, arguably one of the major features of psychopaths is that they are in some ways people who don't have the motivation confusion but are able to lie well.

It seems to me from this conversation that our views are not as far apart as they initially seemed, and in so far as they disagree, you seem to have made good points about the presence of sexual selection being likely to have certain results. It seems that remaining disagreement is to a large extent based on background intuitions and vague words like the difference between "unlikely" and "very unlikely". So, unless one or both of us tries to be a lot more precise, or unless we encounter some non-human evolved HLIs, it isn't obvious to me what to say at this point. You've caused me to update my estimate for how close I should expect evolved HLIs to mentally resemble humans in the direction of expecting them to be more similar but I'm not sure how much I should update in that direction.

Robin Hanson has an interesting argument here.

To flesh out his statistical argument as it applied to this case: there are certain general trends (what you call universals) across most animals, and most human societies; while there exist species (and cultures) that are exceptions to many of them, there's are reason they are exceptions.

To take one of your examples, while it may be short term beneficial for a species to reproduce completely without genetic recombination, it greatly limits potential future evolution since it's no longer possible for beneficial mutations from different individuals to combine.

Edit: See also Eliezer's post evolving to extinction.

[-]FAWS-40

They might lack our bizarre confusion between being able to act according to ones preferences and ones future actions being uncertain in some sense. I don't just mean the libertarian free will nonsense, even someone very intelligent and long past that particular idea like Vladimir Nesov still seems to think that being able to decide requires logical uncertainty about ones decision.