3 ms·

Sooo... can we breed octopi for intelligence? How far could that rise in say a hundred years? and I just found this...: http://danielkeating.blogspot.com.au/20

by D_Alex 11y ago

Sooo... can we breed octopi for intelligence? How far could that rise in say a hundred years? and I just found this...: http://danielkeating.blogspot.com.au/2012/01/could-we-breed-dogs-as-smart-as-people.html http://danielkeating.blogspot.com.au/2012/01/could-we-breed-...

joshmarlow 11y ago
The idea of breeding cephalopods for increased intelligence is touched on in the (great) novel, "Manifold: Time". Not cephalopods specifically, but the idea of breeding smarter animals is touched on by David Brin in his "Uplift" novels. I've often thought about how fascinating it would be if some organization took on the long-term task of breeding for intelligence. Perhaps it would require a well-funded non-profit to provide the resources. Could it be done more open-source styled, with standardized tests, individuals breeding animals and submitting gene sequences for analysis (is there a 23andme for animals)? How interesting would it be to (ultimately) talk to a gorilla or octopus philosopher?
- cgriswald 11y ago
  Probably very interesting. Being bred for human style intelligence, it would have similar capabilities to our own in terms of cognitive capacities. Yet its metaphors for interacting with the world would all be based on the body of an octopus. It would have a tenticle in both worlds. Some things would overlap, like big meaning important, sharp meaning dangerous or cutting. Others might not, for instance any metaphor having to do with wetness would probably have a different meaning for the octopus whose home is the water than it does for those of us on land.
- WalterBright 11y ago
  We breed dogs for all kinds of absurd features. Why not try and breed them for intelligence?
  - MrJagil 11y ago
    I believe we do, labradors and such for sniffing out drugs and explosives or guiding blind people.
- Pyxl101 11y ago
  At what point do you have to stop the experiment because the species has become too smart? A human-level intelligence that can make arguments in its own defense has a good likelihood of being recognized as having rights by a court.
- HCIdivision17 11y ago
  That novel completely ruined squids and asteroid mining for me, because I immediately think about it. Whenever I come across a group studying cephalopod intelligence, I am waiting for someone to say they developed a computer interface that they glom to and have one operating a motorized tank.
thaumasiotes 11y ago
Of course; you can breed for anything. But intelligence wouldn't be very interesting -- what you'd really want is to get them to communicate.
- TheOtherHobbes 11y ago
  You could try ASCII - they should have no problem with octal. It's hard to measure IQ without a social context. Humans are unique in that most of our collective learning is shareable and externalised. If you took a human child and kept it completely isolated from all external teaching and human contact until after puberty, it's not obvious that it would appear that much more intelligent than an octopus. Conversely, if you could give an octopus the genes for a much longer lifespan and less need to be solitary, I don't know what you'd get - but it might be scary smart.
  - sirsar 11y ago
    > If you took a human child and kept it completely isolated from all external teaching and human contact until after puberty, it's not obvious that it would appear that much more intelligent than an octopus. Obviously we cannot deliberately test this hypothesis, but there have been a few cases of children raised in isolation who failed to ever attain an adult level of language [0]. In fact, if you measure by number of acquired words alone (a dubious metric), their capacity for learning language was at the same order of magnitude as gorillas [1] (the validity of gorilla language studies remains controversial). To be taken with a rock of salt, of course, but it shows that development is far from guaranteed. This makes determining the intelligence potential of a species very difficult: what is the optimal environment for baby octopi to learn to communicate? [0] https://en.wikipedia.org/wiki/Language_deprivation#Cases_of_language_deprivation https://en.wikipedia.org/wiki/Language_deprivation#Cases_of_... [1] https://en.wikipedia.org/wiki/Koko_(gorilla) https://en.wikipedia.org/wiki/Koko_(gorilla)
kanzure 11y ago
> Sooo... can we breed octopi for intelligence? How far could that rise in say a hundred years? Who has time to wait 100 years? Intelligence is not really correlated with brain size, but I'd start by breeding octopodes and then selecting the octopus eggs that seem to be inhabited by octopus with larger-than-average brains. Their eggs are transparent and you can just look, while tossing the ones that you don't want to hatch. Next I would iterate on this for a few generations, meanwhile selecting for octopus problem solving and communication.
- sp332 11y ago
  100 years is not a long time. And physical brain size has a low correlation with intelligence.
- slowmovintarget 11y ago
  But the article points out that not all of their neural structure is in their discrete brain case. Part of the machinery that drives their intelligence seems to be in their limbs.
  - kanzure 11y ago
    > Part of the machinery that drives their intelligence seems to be in their limbs. I wouldn't worry about that; their limbs are welcome along for the ride.
- gwern 11y ago
  I think that might quickly fail. I know you've already seen the guppy paper where they did just that with mixed results, and that was only a handful of generations. Brain size is not that great a proxy for intelligence.
  - kanzure 11y ago
    > Brain size is not that great a proxy for intelligence. Oh certainly, agreed. But larger brains can be easier to work with, such as for slicing studies where you examine the different areas of the brain to look for evidence of changes that you might want to keep for the next generation. Smaller brains have less matter to work with, and therefore have less of a chance of developing novel features that strike your fancy while you attempt to select the ones that might have a higher probability of developing impressive cognitive ability. Having said this, I am very interested in trying to take something already rather clever (like a dog) and seeing how small we can get the brain before it loses all of its interesting behavior. For whatever reason society has already developed a fascination with really small dogs with small heads anyway.....
shkkmo 11y ago
It seems like the thing you need to breed for is not intelligence, but longevity, social behavior, and a drive towards communication.
phdfreddied 11y ago
Octopodes, not Octopi. Octopuses if you must. https://en.wikipedia.org/wiki/Octopus#Etymology_and_pluralization https://en.wikipedia.org/wiki/Octopus#Etymology_and_pluraliz...

by D_Alex 11y ago

joshmarlow 11y ago

The idea of breeding cephalopods for increased intelligence is touched on in the (great) novel, "Manifold: Time". Not cephalopods specifically, but the idea of breeding smarter animals is touched on by David Brin in his "Uplift" novels. I've often thought about how fascinating it would be if some organization took on the long-term task of breeding for intelligence. Perhaps it would require a well-funded non-profit to provide the resources. Could it be done more open-source styled, with standardized tests, individuals breeding animals and submitting gene sequences for analysis (is there a 23andme for animals)? How interesting would it be to (ultimately) talk to a gorilla or octopus philosopher?

cgriswald 11y ago
Probably very interesting. Being bred for human style intelligence, it would have similar capabilities to our own in terms of cognitive capacities. Yet its metaphors for interacting with the world would all be based on the body of an octopus. It would have a tenticle in both worlds. Some things would overlap, like big meaning important, sharp meaning dangerous or cutting. Others might not, for instance any metaphor having to do with wetness would probably have a different meaning for the octopus whose home is the water than it does for those of us on land.
WalterBright 11y ago
We breed dogs for all kinds of absurd features. Why not try and breed them for intelligence?
- MrJagil 11y ago
  I believe we do, labradors and such for sniffing out drugs and explosives or guiding blind people.
Pyxl101 11y ago
At what point do you have to stop the experiment because the species has become too smart? A human-level intelligence that can make arguments in its own defense has a good likelihood of being recognized as having rights by a court.
HCIdivision17 11y ago
That novel completely ruined squids and asteroid mining for me, because I immediately think about it. Whenever I come across a group studying cephalopod intelligence, I am waiting for someone to say they developed a computer interface that they glom to and have one operating a motorized tank.

thaumasiotes 11y ago

Of course; you can breed for anything. But intelligence wouldn't be very interesting -- what you'd really want is to get them to communicate.

TheOtherHobbes 11y ago
You could try ASCII - they should have no problem with octal. It's hard to measure IQ without a social context. Humans are unique in that most of our collective learning is shareable and externalised. If you took a human child and kept it completely isolated from all external teaching and human contact until after puberty, it's not obvious that it would appear that much more intelligent than an octopus. Conversely, if you could give an octopus the genes for a much longer lifespan and less need to be solitary, I don't know what you'd get - but it might be scary smart.
- sirsar 11y ago
  > If you took a human child and kept it completely isolated from all external teaching and human contact until after puberty, it's not obvious that it would appear that much more intelligent than an octopus. Obviously we cannot deliberately test this hypothesis, but there have been a few cases of children raised in isolation who failed to ever attain an adult level of language [0]. In fact, if you measure by number of acquired words alone (a dubious metric), their capacity for learning language was at the same order of magnitude as gorillas [1] (the validity of gorilla language studies remains controversial). To be taken with a rock of salt, of course, but it shows that development is far from guaranteed. This makes determining the intelligence potential of a species very difficult: what is the optimal environment for baby octopi to learn to communicate? [0] https://en.wikipedia.org/wiki/Language_deprivation#Cases_of_language_deprivation https://en.wikipedia.org/wiki/Language_deprivation#Cases_of_... [1] https://en.wikipedia.org/wiki/Koko_(gorilla) https://en.wikipedia.org/wiki/Koko_(gorilla)

kanzure 11y ago

> Sooo... can we breed octopi for intelligence? How far could that rise in say a hundred years? Who has time to wait 100 years? Intelligence is not really correlated with brain size, but I'd start by breeding octopodes and then selecting the octopus eggs that seem to be inhabited by octopus with larger-than-average brains. Their eggs are transparent and you can just look, while tossing the ones that you don't want to hatch. Next I would iterate on this for a few generations, meanwhile selecting for octopus problem solving and communication.

sp332 11y ago
100 years is not a long time. And physical brain size has a low correlation with intelligence.
slowmovintarget 11y ago
But the article points out that not all of their neural structure is in their discrete brain case. Part of the machinery that drives their intelligence seems to be in their limbs.
- kanzure 11y ago
  > Part of the machinery that drives their intelligence seems to be in their limbs. I wouldn't worry about that; their limbs are welcome along for the ride.
gwern 11y ago
I think that might quickly fail. I know you've already seen the guppy paper where they did just that with mixed results, and that was only a handful of generations. Brain size is not that great a proxy for intelligence.
- kanzure 11y ago
  > Brain size is not that great a proxy for intelligence. Oh certainly, agreed. But larger brains can be easier to work with, such as for slicing studies where you examine the different areas of the brain to look for evidence of changes that you might want to keep for the next generation. Smaller brains have less matter to work with, and therefore have less of a chance of developing novel features that strike your fancy while you attempt to select the ones that might have a higher probability of developing impressive cognitive ability. Having said this, I am very interested in trying to take something already rather clever (like a dog) and seeing how small we can get the brain before it loses all of its interesting behavior. For whatever reason society has already developed a fascination with really small dogs with small heads anyway.....

shkkmo 11y ago

It seems like the thing you need to breed for is not intelligence, but longevity, social behavior, and a drive towards communication.

phdfreddied 11y ago

Octopodes, not Octopi. Octopuses if you must. https://en.wikipedia.org/wiki/Octopus#Etymology_and_pluralization https://en.wikipedia.org/wiki/Octopus#Etymology_and_pluraliz...