A partial index of discussion notes is in
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/AREADME.html
http://www.cs.bham.ac.uk/~axs/pics/feeder/squirrel-frame-right
What does that tell us about squirrel consciousness?
There are many more online videos showing competitions between squirrels and bird-feeders. Until recently the winner was usually a squirrel. Humans seem to have improved their ability to design anti-squirrel features into bird feeders.
One or more slugs are then found in the tray, apparently feeding on the (partly crushed) nuts. Sometimes the trails still exist, but the slugs are not there, presumably because they have gone down the window .... or been eaten by birds!
What does that tell us about slug consciousness?
Some pictures of slugs are available here:
https://www.shutterstock.com/search/slug
Of course it moves through a continually changing local space in a way that produces a path through a non-local space. But I have no reason to believe the slug knows that that is what it is doing.
A squirrel would not be able to survive if it were restricted to such a strategy. E.g. in order to get to food (or its nesting site in our loft where baby squirrels have been left) a squirrel frequently has to select a route that is not constantly bringing it closer to its goal.
A squirrel looking up from our patio to the tempting nuts has a vast collection of possible actions, including sensing actions, available to it. These actions can be assembled into action sequences, but instead of trying increasingly complex combinations of available actions (some of which would be fatal, others utterly useless) a squirrel can often be observed to look around to identify intermediate locations on accessible structures and then select one to try. For instance, the squirrel may look at the bench and table visible in some of the pictures, and rule them out even though they would reduce the height gap between it and the nuts.
In a different situation a squirrel will move away from the target and climb up a structure that takes it to a greater height than the target, at a greater distance than its present distance. Having reached that intermediate target it leaps across to the final target, covering a large horizontal distance while its height is reducing, as in this video, which will not surprise anyone familiar with (grey?) squirrels: https://www.youtube.com/watch?v=HyomSW3PMQM
Notice how, after its initial jump fails, it looks around, apparently in order to select a new route. (There are many online videos showing squirrels battling bird-feeders.)
The intelligence of Portia spiders may also illustrate the role of a meta-configured genome insofar as no standard "bottom-up" learning mechanism could produce the route planning capabilities of Portia.[Explain why not!] Neither could squirrel intelligence and Portia intelligence all be due to some general purpose inductive learning machine, since the Portia abilities seem to be unique to that species (and close relatives?).
Similar remarks can be made about human toddlers (at various stages of development), as illustrated in a discussion of "toddler theorems". Although squirrels and human toddlers may both be described as far more intelligent than slugs and snails, the abstract question question whether a particular squirrel is more or less intelligent than a particular toddler makes no sense, since there are spatial configurations and problems that a squirrel may be able to deal with that a certain toddler cannot and vice versa. In part this may be due to differences in physical structures available for exercising intelligence (e.g. muscular/skeletal structure -- including structures of jaws, arms, wrists, hands and fingers, offering different opportunities for manipulation of physical objects, or for self-propulsion, along a horizontal surface, or along a tree branch, or between disconnected spatial structures (e.g. leaping between trees).
Different information processing (i.e. cognitive) mechanisms and capabilities are required for these competences, though there are some overlaps. According to the Meta-configured Genome (MCG) theory, in some species the mechanisms are not fully specified in the genome, and not acquired by interactively training/adapting a general ability until it match a specific environment. Instead the theory states that information collected and stored during certain interactions with the environment can be used to fix parameters for genetically partly specified mechanisms that are produced at a later stage of gene expression.
So instead of late expressed, late evolved, capabilities making use of a process of training a mechanism to match the environment, the mechanism is specified with gaps for parameters to be derived from results of earlier interaction with the environment. So the same genome can produce different capabilities in different environments not because of a common training process, but because results of a general data-collection mechanism at an earlier stage of development can provide parameters for genetically specified mechanisms produced at a later stage. This will depend on the earlier learning process having some use for the individual organisms when first evolved, perhaps a social use, e.g. when very young infants learn to imitate syllable sounds produced by others, perhaps rewarded only by simple social interactions with others, not by communicated semantic content. What is acquired at that stage provides a platform of components for building/understanding relatively complex semantic communications at a later stage of genome expression.
Here's an extract from another paper:
More complex internal meaning structures are required for cognitive functions based on information contents that can vary in structure and complexity, like the Portia spider's ability to study a scene for about 20 minutes and then climb a branching structure to reach a position above its prey, and then drop down for its meal (Tarsitano, 2006). This requires an initial process of information collection and storage in a scene-specific structured form that later allows a pre-computed branching path to be followed even though the prey is not always visible during the process, and portions of the scene that are visible keep changing as the spider moves. Portia is clearly conscious of much of the environment, during and after plan-construction. As far as I know, nobody understands in detail what the information processing mechanisms are that enable the spider to take in scene structures and construct a usable 3-D route plan, though we can analyse the computational requirements on the basis of half a century of AI experience.
M. Tarsitano, (2006), Route selection by a jumping spider (Portia labiata) during the locomotory phase of a detour, Animal Behaviour vol 72, Issue 6, pp. 1437--1442, http://dx.doi.org/10.1016/j.anbehav.2006.05.007Extracted from: Aaron Sloman, (2016), Natural Vision and Mathematics: Seeing Impossibilities, in Second Workshop on: Bridging the Gap between Human and Automated Reasoning, at IJCAI 2016, pp.86--101, Eds. Ulrich Furbach and Claudia Schon, July, 9, New York, http://ceur-ws.org/Vol-1651/ Available here:
http://www.cs.bham.ac.uk/research/projects/cogaff/Sloman-bridging-gap-2016.pdf
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/triangle-sum.html
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/trisect.html
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/triangle-theorem.html
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/impossible.html
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/torus.html
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/shirt.html
There are also toddler-theorems:
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/toddler-theorems.html
Another challenge for automated reasoning systems -- discover/invent a 3-D
ontology in order to explain/understand sensed 2-D phenomena:
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/nature-nurture-cube.html
Maintained by
Aaron Sloman
School of Computer Science
The University of Birmingham