6: Introducing The Rat Brain Robot
The author considers an experiment done at the University of Reading (2008) to have gotten far less attention than in merits: Ben Whalley has taken a cluster of about 300K brain cells from a rat and wired it up to a robot, in such a way that the "rat brain" controls and receives feedback from the machine.
While there's a bit of excitement in the press (the author cites a BCC report), progress is slow and the results are not quite sensational enough to spin up the media, but from a science perspective, the integration of organic components to a mechanical system, examining the most fundamental elements of memory, emotion, and decision-making.
The author provides a (broken) link to a video of the robot rat: the brain itself is kept in a cabinet, and is connected wirelessly to a small robot, a mobile unit with a camera, which navigated around a maze avoiding obstacles. It's a fairly standard rat-in-a-maze experiment, such as you might see at any junior-high science fair - the key difference being that the "rat," being mechanical, does not have any physical sensations, stimulus or sensation, to motivate it. There is no cheese a t the goal to be smelled, nor any motivation to eat it even if there were.
Brain Sizes
The robot rat is based on a cluster of 300K neurons, perhaps 2% of the total number that exist in the natural animal. But even this is more than sufficient to the task, considering that it's still three times as many as a fruity fly, and a fruit fly does some amazing things: it is able to fly, after all, not to mention basic survival skills (find food, avoid danger, mate, etc.)
There's then a list of animals and the number of neurons they have:
- 0 - Sponge
- 10K - Leech
- 100K - Fruit fly, Lobster
- 250K - Ant
- 1M - Cockroach, Honey Bee
- 4M - Mouse*
- 15M -Frog
- 20M - Rat*
- 160M - Dog*
- 300M - Cat*
- 6B - Chimpanzee*, Dolphin*
- 10B - Human*
- 20B - Elephant
- 30B - Pilot Whale*
It's noted that the number above counts only those in the cerebral cortex of some animals (asterisks). It other species, it's for the entire body. (EN: I did some additional digging and updated/augmented the author's original list and grouped a few items that were close. Also worth noting the estimates I saw for humans range from 80 to 120 billion, not the 10 billion the author suggests.)
Animal Intelligence and Human Intelligence
By the table above, it should be clear that the gross number of neurons is related to intelligence, but does not drive intelligence: you could not glob together the neurons of 11,000 cockroaches to "make" a brain that is as intelligent as a human being; nor have pilot whales established great civilizations under the sea.
A certain number of neurons are necessary to basic survival functions autonomously: we do not need to "think" about each breath and heartbeat or undertake a mental effort to keep these systems running. These homeostatic functions "just happen" without deliberate thought. Sensation, too, is autonomous. We do not, and cannot, choose what we hear, see, smell, taste, or feel - though we may concentrate intently on one thing and become oblivious to others.
One of the key differences between animal and human brains is the autonomy of responses. Animals have instincts, which cause them to react to stimulus in a consistent manner, without evaluating the sense-data or considering alternative courses of actions. Humans can act in the same manner, when they allow their emotions to take control and refuse to think, but also have the ability to apply reason and consider their experience - and because they do so, they can react in a different manner, and the outcome may be better or worse than animal instinct.
(EN: Calls to mind something I once heard in a documentary about survival. A human being can survive where animals cannot. The specific example was that a dog will spit out anything that tastes unusual, even if it is nourishment, and will starve to death. A human will want to spit it out, but will overcome his emotional reaction and be able to consume it, learn that it is edible, and thus survive.)
The decision-making process in lower orders of animal is not the same as planning: they have an infinite number of possible actions they may take at any time, and "choose" a specific course of action, but they are driven by pre-programmed instinct to do what is best (avoid pain and seek pleasure) at any given time.
Higher orders of animals develop planning skills by way of experience. Even at the level of rodents, a creature is able to remember when a choice caused them pain or granted a reward, associate the experience of pain or pleasure to sense-data, and add this to their instinctual pre-programming. It's by this means that a rat learns to navigate a maze to find food or a cat knows the sound of an electric can opener.
Aside of the instincts that drive animals to act for their individual survival, there are also instincts that control behavior in a group of similar creatures. Insects such as ants and bees have highly developed social instincts. Mammals have significantly less developed social instincts, but those that travel in herds or packs have some hard-wired social instincts that can interfere with their individual survival instincts, enabling them to work together to improve their collective survival.
This leads to specialization of creatures that goes beyond the physiological: some are leaders, some are followers, some are workers, some are soldiers, and so on. Another experiment at Reading demonstrated that, by altering the genetic structure can change the "adventurousness" trait of a rat, evidence that personality is, at least to some degree, physiological.
The author stresses that his exploration here is minimal - there are not the only systems in the brain, and decision-making is highly complex even in some species of animal - they are merely examples that relate most closely to the topic of this book: sensation, emotional response, and behavior.
Movement and Motivation
In a general sense, movement is what differentiates animals from plants: plants are located where their seed fell. They thrive or die, or fail to grow at all, based on environmental conditions - they cannot flee from danger or move to a more attractive location.
Animals survive by moving about, both in terms of immediate survival (flee danger, find food) and their long-term survival (migrate from an area that lacks food, water, climate, etc. to one that is more suitable.)
(EN: The author skips the last step, which is human survival's difference from animals, particularly for long-term survival: we plant crops, dig wells, build dwellings, etc. to adapt the environment rather than moving ourselves. This is highly significant to marketing.)
An interesting case is the "sea squirt", an animal that hatches from an egg, moves to a location, and anchors itself. Once anchored, it no longer needs to make decisions, and digests its own brain.
The mere ability to move is separate from motivation - the desire to exercise that ability, and the decision-making process that tells us where to go and where to avoid.
Even primitive animals move: amoebas will twitch their muscles in response to heat or cold and move in the opposite direction, and the pupa inside a jumping bean will move to avoid heat. Higher orders of animals are more complex, as they begin to exercise more complex reactions.
In its most basic sense, this ability depends upon our choice of a goal, and our observations of options along the way. In the rat-maze experiment, it is reasoned that the animal knows roughly where the cheese is (by scent) and then navigates a path between barriers (where it cannot go) and spaces (where it can).
We learn the basics fairly quickly - that we cannot walk through walls, or walk on air, or walk on water. We fumble at a solution - that a maze can be navigated by keeping your right hand against one wall. We discover shortcuts - observing a path that can be taken that gets to the goal more quickly.
More importantly, once a problem is solved, we remember to repeat the same solution the next time we need to solve it, or discover more efficient ways of solving it, such as turning right, rather than left, at a certain point in a maze to cut off a loop.
And so, the incentive to act - motivation - is the topic that is of greatest interest to marketers. We know that customers purchase, but getting them to purchase our brand, as opposed to another one, and as opposed to doing without, is in understanding the motivation that drives action.