The bird brian

Science General Neuroscience
1

I just saw a Instagram video of a crow drinking water out of a bottle. When his beak could no longer reacher the water, this crow grabbed some pebbles and dropped them in the bottle to cause the water to rise so he could get more water. I’ve never seen intelligence like that with my dogs lol so it made me curious about the “bird brain”.

Anyway reading some short articles I read that scientists think their could be similar structures that perform “intelligence” like the neocortex does in mammals.

So I’m wondering if studying the birds “neocortex variant “ could help in the study of come up with true intelligence. Surely the bird brain structure is doing the same stuff the mammal brain does. Would it be worth wild to compare and contrast; see how the bird brain does it differently, maybe give new insight on how the neocortex does it. Or maybe their is a similar “circuit “ that we can simulate and make more of to achieve advance intelligence beyond what the smartest bird could do? Just thinking out loud I guess, any thoughts or info?

1 Like
2

This was originally discussed (think ancient Greece) in an Aesop fable. You might find the following provocative, but do not confuse self-awareness with sentience.

3

There was a similar discussion to this here: Cognition without Cortex

I’m regularly impressed by what non-mammals can do!

4

Keep in mind (no pun intended) that all the magic happens because of language.

5

Who names a bird Brian anyway?

I mean, I grew up with a canary called Tweety and a chickling called Calimero.

But… Brian?

6

Spot on. There is an evolved brain function that provides complex abilities we call intelligence. Mammals have it and birds have it and the bigger it is, the smarter the animal is. Primates are top of the mammal class, corvids are top of the avian class, but other birds are smart too (magpies, currawongs, cockatoos, etc).

I don’t give a rats about neuroanatomy and all the little thalamus/hippocamplus/other complex evolved structures. I don’t see us ever working out how they do what they do.

The elephant in the room and by far the most plausible target for our efforts is the neocortex in mammals and the pallium in birds. The structure is relatively simple, the level of function is high and the more you have the smarter you are.

7

Birds (corvids included) do not have their pallium arranged in 6-layer macrocolumns, right? So then it would stand to reason that the core algorithm behind biological intelligence cannot be specific to macrocolumn structure.

1 Like
8

Actually the language happens because of the magic. Without intelligence first, language proper cannot happen. To wit (no pun intended): voice actuated morons like Alexa and my car.

1 Like
9

True. My view is that the columnar structure might convey an advantage in ‘scaling up’: an adaptation resulting in more columns results in more intelligence sooner.

But the key point is that we have the mammalian cortex and the avian nidopallium evolving separately but producing a similar final result. That must be helpful to our discovering how. See Nidopallium - an overview | ScienceDirect Topics which concludes:

> Thus, nonhomologous brain areas converged over the course of 300 million years into mammalian and avian prefrontal structures that serve highly similar functions. In doing so, both areas gained the ability to generate the same cognitive functions using similar cellular properties.

10

I could not agree more, except I would say that ‘the core algorithm behind human intelligence cannot be specific to macrocolumn structure.’

The human brain develops a self-organizing OS called Consciousness. This is not reactivity, awareness, or attention even though it operates with them and often in spite of them. C gives rise to the astonishing repertoire of human civilization and technology. A bird figuring out how to raise the water level in a flask so it can drink is impressive, but certainly not C.

1 Like
11

I strongly suspect that patterns of connections is very important to configuring function.

Given some basic local functions, connecting them is as important to the final global functions. Three or six layers is perhaps not as important as that the three layer parts in common do the same things.

Different critters brains look very different and researchers make it worse by calling the same parts different names but the connections between the parts follow some common general schemes.

I am (slowly) working to understand how this ridiculously complicated mess is connected.

1 Like
12

Yes. In general neuroanatomy is a bit like taking photos of printed circuit boards and hoping it makes sense when you don’t know what the chips do.

I went to school with images like this:
Cajal_cortex_cropped_drawings

There was a strong suggestion that brain neurons were sparse, with lots of glial cells and similar in between. But the reality seems to be that the space is jammed full with dendrites and synapses and that brain neurons are hyper-connected. Billions of neurons but trillions of connections.

And that’s a problem for neuron-oriented models like ANN, or connectome etc that try to understand individual connections. Anatomy getting in the way again.

IMO SDR is the right way to think about sets of connections/synaptic activations. HTM is a step along the path to understanding, but only a first small step so far.

1 Like
13

The more I study (or try to and fail in various ways) the biological architecture of the brain the common human failing seems to be that we have a tendancy to apply an understanding of the complexity, which fits within a visual or linguistic type framework. Once you do down the path of more than 4 dimensions, the process goes into an abstract realm within thought and may then become a problem to deal with (without the derivative abstract throught process we call maths reverting the problem to 2D space - on a bit of paper).

One thing that I’m not sure about is do we have a tendancy to collapse “thought” to words by virtue of say the arcuate fascillius and therefore limit our ability or degree of abstract thinking ? Bear with me on this as it does get more abstract… The theory here is that the AF helps regulates “thought” by virtue that we need to be able to externalise in communication, to others or create serial actions, any such “valid” throught process (in a caveman realm, or pre-internet days…). This internal mechanism then acts like a combined speed governor and attention mechanism and may be particular to humans in that way as it becomes more of a necessity as complexity increases.

The brain, being massively interconnected in a hierarchical mesh type structure, by default processes waves of activity in parallel, much like throwing a hand full of gravel into a pond at an angle… the stones falling in the water ripple through the further away they land. The stones being akin to say sensory input or distal interconnected dendritic branches, phased in time in part by the hippocampus (the DG and HC being a particular temporal buffer and phase management structure). The key here is the massively parallel type process going on still need the right sequential structure activation.

Where birds come into all this… is that the buffer and phasing of “thought” is limited in dimensionality, which therefore limits the ability of complexity in understanding. Think of it along the lines of using an alphabet with only say 6 characters and a maximum word grouping of say 3 words. Dimensional constraint in a temporal bound limit complexity.

Language (to me) very clearly shows this buffering and phasing aspect of sensory input streams, where language shows a very particular type of structure that does not sit well with any existing formal deffinitions. This buffering and phasing is maybe the key limit to what goes on. I have modelled part of this activity and it does seems to fit a part, but still figuring out if or how the part fits overall.

The only difference humans may actually have is that thier “buffer” just has more space, nothing else (maybe the AF plays a particualr role in what we call consiousness). The dregree of complexity can only arrive as a derivative of the number of bits available and the bounds in which those bits can apply.

Birds just have a limited buffer, they have the right bits, just not enough of them…

Where it get’s really abstract for me is that if we consider that the human brain is dimensionally constrained by probablistic dendritic proximity what happens with an artificial emulation that does not have a 3D physical constraint…

Maybe just my abstract way of seeing the world… just some thoughts to hopefully create more ideas…

1 Like
14

No, I don’t find neuroanatomy helpful here - you can omit the AF and the rest. But yes, Intelligence precedes and is not dependent on language and works fine without. And no, we don’t need more dimensions.

At the simplest level avian intelligence is constrained by brain size. Avain brains have the hyper-connections, if not with columns, but their <1W cannot match our 20W.

If we understood and could replicate the hyper-connected brain and its algorithms in silicon and software we would gain a x1000000 speed up.

Asimov asked: is there a god? Answer: yes, now there is a god.

1 Like