How complicated can genetically coded behaviors be?

[Tvtyrant]

This is a question that comes out of animals like ants, and sci-fi creatures such as Orks. Many (most?) animals rely on instincts and coded behavioral patterns which are somehow imprinted into their DNA (some songbirds when raised in isolation know how to sing their species' songs).

Is it known how DNA can imprint instinctual functions onto an animal? What are the limits to this kind of genetic memory/behavior?

[halfeye]

This is a question that comes out of animals like ants, and sci-fi creatures such as Orks. Many (most?) animals rely on instincts and coded behavioral patterns which are somehow imprinted into their DNA (some songbirds when raised in isolation know how to sing their species' songs).

Is it known how DNA can imprint instinctual functions onto an animal? What are the limits to this kind of genetic memory/behavior?

How much DNA do you have available? I don't know that the mechanism is known in detail, but I'm pretty sure the answer is going to be in the region of huge amounts.

[Tvtyrant]

How much DNA do you have available? I don't know that the mechanism is known in detail, but I'm pretty sure the answer is going to be in the region of huge amounts.

Is this a limitation in nature? Some organisms (https://en.wikipedia.org/wiki/Polychaos_dubium)have huge amounts, so assuming in that region.

I guess my question is, could a creature with developed technology gain it through genetic coding instead of through memory learning?

[Grinner]

I guess my question is, could a creature with developed technology gain it through genetic coding instead of through memory learning?

Sure? Things like songbird songs exist as neural chains. There are going to be limits on behavior based on the size of the genome, because while individual genes can serve multiple purposes through alternative splicing, they can still only do so much. But that's still a lot.

Ants, and most swarm intelligence agents in general, don't actually have much cognitive capacity on their own, but they have behaviors which combine to solve problems through emergence. For example, I remember hearing about experiments done with termites. A single termite can be placed in an enclosed area full of tiny pellets, and they'll just flounder to accomplish anything. If a couple more termites are added, they collectively suddenly start building definite structures.

Getting back to songbirds, the neural chains in their brains reflexively activate their muscles in certain patterns to create their songs. Your genetically-encoded technology might be similar, but in this case, the neural chain might activate other neurons in such a way that the organism gets an idea instead of producing a song. But this is also dependent on having other neural structures in reliable places. It's not impossible, but how feasible it is depends on a number of factors.

To get back to the original question, no, I have no idea how genes produces proteins which in turn sculpt neural structures.

[Tvtyrant]

Sure? Things like songbird songs exist as neural chains. There are going to be limits on behavior based on the size of the genome, because while individual genes can serve multiple purposes through alternative splicing, they can still only do so much. But that's still a lot.

Ants, and most swarm intelligence agents in general, don't actually have much cognitive capacity on their own, but they have behaviors which combine to solve problems through emergence. For example, I remember hearing about experiments done with termites. A single termite can be placed in an enclosed area full of tiny pellets, and they'll just flounder to accomplish anything. If a couple more termites are added, they collectively suddenly start building definite structures.

Getting back to songbirds, the neural chains in their brains reflexively activate their muscles in certain patterns to create their songs. Your genetically-encoded technology might be similar, but in this case, the neural chain might activate other neurons in such a way that the organism gets an idea instead of producing a song. But this is also dependent on having other neural structures in reliable places. It's not impossible, but how feasible it is depends on a number of factors.

To get back to the original question, no, I have no idea how genes produces proteins which in turn sculpt neural structures.

Thank you! I know apes have relatively few programmed responses compared to other animals (like having to learn to swim) but Dolphins and Crows are both problem solving and have lots so for Sci-Fi they seem not incompatible.

I can also now phrase the question correctly :D

[The Great Wyrm]

Think of the difference between an ASIC and a FPGA. You can make both do the same thing, but there are advantages and disadvantages to each.

[Rockphed]

Think of the difference between an ASIC and a FPGA. You can make both do the same thing, but there are advantages and disadvantages to each.

Technically, fpgas are the middle ground between cpus and asics. They aren't quite as fast as asics and aren't quite as versatile as cpus, but are much, much cheaper to deploy than asics, and much, much faster than cpus. I'm not sure this is germane since there are neither cpu- nor asic-analogue brains.

[Phase]

Virtually all behavior in nature, from bacteria to archaea to eukarya, is a relatively simple call-response system. There's a stimulus such as a chemical gradient, this interfaces with receptors that trigger a change in the cellular skeleton, and the organism moves in response to the stimulus. This becomes rather more complicated in multicellular organisms, but mostly because of how the structures involved become more intricate.

We don't fully understand the molecular basis for memory, which is the biggest issue with this idea, but even if we did, the molecular basis for a particular memory is even further out of reach.

An interesting point you bring up is the instinctual behaviors of eusocial insects like ants. What's astonishing is that a worker ant and a soldier ant born to the same queen are virtually identical on a genetic level. Their deviant behaviors and physiologies are determined by methods of DNA regulation that differ between them. A study I read recently was even able to make one caste of ant behave like another by feeding it a drug that inhibited this one protein. In the end, the extreme diversity of behaviors you see in, say, birds or mammals is just generated more efficiently through learning and memory.