Biological computation: By evolution, or design?
https://reasonandscience.catsboard.com/t3073-biological-computation-by-evolution-or-design
Artificial Intelligence performing algorithmic calculations are always the product of human engineering. It has never been demonstrated that a machine can become a mind. A computer can perform many different procedures; process algorithms — a set of distinct, well-defined steps. Selection sort accomplishes a task with one basic operation that, when performed over and over, completes the whole task. It also has a well-defined start state and end state, which can be referred to as its input and output. Algorithms have a well-defined set of steps for transforming input to output. So anyone who executes an algorithm will perform the same steps, and an algorithm’s output for a given input will be the same every time it is executed An algorithm doesn’t necessarily have to involve repetition, but any task performed on a large set of data usually will use such repeated steps, known as “loops.”
Take, for example, books that have titles composed of known characters, allowing for alphabetization; the shelf has an ordering (beginning to end, or left to right); the books are objects that can fit onto the shelf and be moved about; and so on. Only these few properties are relevant for the purposes of the algorithm. The sorter needs to know nothing about the full nature of a book in order to execute the algorithm — you need only have knowledge of shelf positions, titles, and how titles are ordered relative to one another. This abstraction is useful because the objects involved in the algorithm can easily be represented by symbols that describe only these relevant properties.
These two forms of abstraction are at the core of what enables the execution of procedures on a computer. At the level of its basic operations, a computer is both extremely fast and exceedingly stupid. The power of the computer derives not from its ability to perform complex operations, but from its ability to perform many simple operations very quickly. Any complex procedure that a computer performs must be reduced to the primitive operations that a computer can execute, which may require many levels at which the procedure is broken down into simpler and still simpler steps.
To do so, you must be able to represent the problem in terms that the computer can understand — but the computer only knows what numbers and memory slots are, not titles or shelves. The solution is to recognize that there is a correspondence between the objects that the computer understands and the relevant properties of the objects involved in the algorithm: for example, numbers and titles both have a definite order. You can use the concepts that the computer understands to symbolize the concepts of your problem: assign each letter to a number so that they will sort in the same way (1 for A, 26 for Z), and write a title as a list of letters represented by numbers; the shelf is in turn represented by a list of titles. You can then reduce the steps of your sorting job into steps at the level of simplicity of the computer’s basic operations. If you do this correctly, the computer can execute your algorithm by performing a series of arithmetical operations. (Of course, getting the computer to physically move your boss’s books is another matter, but it can give you a list ordered the way your boss wanted.)
This is why the computer is sometimes called a “symbol-manipulation machine”: what the computer does is manipulate symbols (numbers) according to instructions that we give it. The physical computer can thus solve problems in the limited sense that we imbue what it does with a meaning that represents our problem.
It is worth dwelling for a moment on the dualistic nature of this symbolism. Symbolic systems have two sides: the abstract concepts of the symbols themselves, and an instantiation of those symbols in a physical object. This dualism means that symbolic systems and their physical instantiations are separable in two important (and mirrored) ways. First, a physical object is independent of the symbols it represents: Any object that represents one set of symbols can also represent countless other symbols. A physical object and a symbolic system are only meaningfully related to each other through a particular encoding scheme. Thus it is only partially correct to say that a computer performs arithmetic calculations. As a physical object, the computer does no such thing — no more than a ball performs physics calculations when you drop it. It is only when we consider the computer through the symbolic system of arithmetic, and the way we have encoded it in the computer, that we can say it performs arithmetic.
A program is independent of the hardware that executes it; it could run just as well on many other pieces of hardware that work in very different ways. But a program is dependent on some physical representation in order to execute. 2
Bat echolocation, signal processing, bioengineering in its finest.
Echolocating animals produce high-frequency sounds through the mouth and use the arrival time, intensity, and frequency content of echo returns to determine the distance, direction, and features of objects in the environment.
By computing the time of reflection of modulates echoes, the bat is able to recognize the object and its distance. Over 1,000 species of bats echolocate with signals produced in their larynges. They use diverse sonar signal designs. Populations of neurons in the bat central nervous system respond selectively to the direction and delay of sonar echoes. In addition, premotor neurons in the bat brain are implicated in the production of sonar calls, along with the movement of the head and ears. Audio-motor circuits, within and across brain regions, lay the neural foundation for acoustic orientation by echolocation in bats. . The active component of echolocation is the animal’s production of sounds that reflect from objects in the environment, and perception is based on information processed from echo returns. The brain does a lot of processing and manipulation of incoming data to build a model of the outside environment.6
Bat's signal strength diminishes with the square of the distance, both going out and coming back. So the outgoing signals must be quite loud for the return echoes to be detectable. But there’s a problem. To pick up the weak return echoes, bat ears have to be extremely sensitive. But such sensitive ears would be wrecked by the loudness of the outgoing signals.
Bats have three distinct design features that must all work together at the same time their voice box, sensitive ears, and unique brains that allow them to understand the signals of echolocation.
So what to do? Bats amazingly turn off their ears during each outgoing chirp and turn them on again to catch each return echo. Ten to 200 times a second!
Question: In order for this system to work, had the mechanism of turning their ears off and on not to be fully developed right from the beginning?
Another problem: Typically there’s a zillion bats around, all creating these echos simultaneously. How can they distinguish their own from all those others? Well, they can, because each has its own distinctive signal. Their brain software masters this too, sorting their own echoes from all the background noise.
Question: How did they "learn" to generate an individual signal in order to distinguish their echoes from background noise? Trial and error?
Today, we all know that bats use echolocation to catch their prey. We take it for granted and see no reason why this fact should be awe-inspiring. Richard Dawkins in The Blind watchmaker page 43:
Donald Griffin tells a story of what happened when he and his colleague Robert Galambos first reported to an astonished conference of zoologists in 1940 their new discovery of the facts of bat echolocation. One distinguished scientist was so indignantly incredulous that he seized Galambos by the shoulders and shook him while complaining that we could not possibly mean such an outrageous suggestion. Radar and sonar were still highly classified developments in military technology, and the notion that bats might do anything even remotely analogous to the latest triumphs of electronic engineering struck most people as not only implausible but emotionally repugnant.
We can only understand it at a level of artificial instrumentation, and mathematical calculations on paper, we find it hard to imagine a little animal doing it in its head. Echo-sounding by bats is just one of the thousands of examples that I could have chosen to make the point about good design. Animals give the appearance of having been designed by a theoretically sophisticated and practically ingenious physicist or engineer, but there is no suggestion that the bats themselves know or understand the theory in the same sense as a physicist understands it. The bat should be thought of as analogous to the police radar trapping instrument, not to the person who designed that instrument. The designer of the police radar speed-meter understood the theory of the Doppler Effect, and expressed this understanding in mathematical equations, explicitly written out on paper. The designer's understanding is embodied in the design of the instrument, but the instrument itself does not understand how it works. The instrument contains electronic components, which are wired up so that they automatically compare two radar frequencies and convert the result into convenient units - miles per hour. The computation involved is complicated, but well within the powers of a small box of modem electronic components wired up in the proper way. Of course, a sophisticated conscious brain did the wiring up (or at least designed the wiring diagram), but no conscious brain is involved in the moment-to-moment working of the box.
Our experience of electronic technology prepares us to accept the idea that unconscious machinery can behave as if it understands complex mathematical ideas. This idea is directly transferable to the workings of living machinery. A bat is a machine, whose internal electronics are so wired up that its wing muscles cause it to home in on insects, as an unconscious guided-missile homes in on an airplane. So far our intuition, derived from technology, is correct. But our experience of technology also prepares us to see the mind of a conscious and purposeful designer in the genesis of sophisticated machinery. One form of the argument makes direct use of the extreme sense of wonder which we all feel when confronted with highly complicated machinery, like the detailed perfection of the echolocation equipment of bats.
Dawkins claim: It is this second intuition that is wrong in the case of living machinery. In the case of living machinery, the 'designer' is unconscious natural selection, the blind watchmaker. It isn't true that the whole perfect work must have been achieved simultaneously. It isn't true that each part is essential for the success of the whole. A simple, rudimentary, half-cocked eye/eat/ echolocation system/cuckoo parasitism system, etc., is better than none at all. How did echolocation get its start? Any animal that can hear at all may hear echoes. Blind humans frequently leam to make use of these echoes. A rudimentary version of such a skill in ancestral mammals would have provided ample raw material for natural selection to build upon, leading up by gradual degrees to the high perfection of bats. Where X is some organ too complex to have arisen by chance in a single step, then according to the theory of evolution by natural selection it must be the case that a fraction of an X is better than no X at all; One hundred and twenty five years on, we know a lot more about animals and plants than Darwin did, and still not a single case is known to me of a complex organ that could not have been formed by numerous successive slight modifications.
Remarkably, besides bats, dolphins, and whales, two birds, the oil-bird and the cave swiftlet also use echolocation, and these two genera have developed the same technology independently of each other. But furthermore, several different kinds of mammals, for instance, shrews, rats, tenrecs and seals seem to use echoes to a small extent, as blind humans do. Acoustic features such as Doppler shift compensation, whispering echolocation, and nasal emission of sound each show multiple convergent origins in bats. 5
Reply: From a computational neuroscience perspective, bats are remarkable because of the very short timescale on which they operate. The barrage of returning sonar echoes from a bat's near-environment lasts approximately 30 milliseconds following a sonar emission with the echo from a specific target lasting, at most, a few milliseconds.
From an engineering standpoint, biosonar systems (e.g. bats and dolphins) have inspired the design of very sophisticated sonar and radar systems that can map distant surfaces and track targets with great precision. Even with powerful mathematical tools and decades of experience, however, our best systems still do not rival the perceptual capabilities of dolphins. Many bats demonstrate incredible aerial agility, flying in complete darkness through branches and caves while hunting evasive insects. These animals perform such tasks in real-time with a total power consumption (including flight) measured in Watts, not hundreds of Watts. In addition to the ability to navigate in complete darkness by echolocation, both bats and dolphins live in very social environments using echolocation in group situations without any obvious problems with interference. All of these capabilities are highly desired by current military programs developing unmanned-aerial vehicles (UAV) especially since many of the target environments are in places where Global Positioning System (GPS) signals are unavailable and obstacle locations are not mapped.”
Bats use echolocation to see objects in front of them. They emit an ultrasonic pulse around 20 kHz (and up to 100 kHz) and then sense the pulses as they reflect off an object and back to the bat. 4
The echolocation process of bats involves three phases to search and capture prey: search phase, approach phase and terminal phase. During the search phase, the bat will start to hunt for prey by emitting the pulse at low rate with frequency around 10Hz. Then, the pulses have to get shorter as the time between the pulse and echo is decreased in order to avoid overlap when the bat spots and gets nearer to the specific prey during the approach phase. In this phase too, pulse emission rate gets steadily increased up to 200 per second since the bat keeps updating the position of the prey. In the terminal phase, the frequency of emitted pulses upsurges more than 200 Hz as the pulse emission rate also starts to accelerate at only a fraction of millisecond long just before the prey is netted
Many bats locate insect prey by emitting ultrasonic pulses and detecting the echoes. By detecting small Doppler shifts ( the change in frequency of a wave in relation to an observer who is moving relative to the wave source ) in the frequency of the returning pulses, the bat's nervous system can discern acoustic 'texture' and so distinguish prey from inanimate objects. Such elegant calculations are based on underlying computational mechanisms.
The Bat echolocation mechanism depends on electrical and mechanical engineering, robotics and control as well as neurological signal processing and sonar signal analysis.
Algorithms, finite sequences of well-defined, computer-implementable instructions, facilitate competent, or even near-optimal, computation; calculations of arithmetical steps that follow a well-defined model. These computations are carried out using biological molecules and cells. In bats, acoustic variables are represented in areas of the bat brain by the electrical 'spikes' emitted by neurons. The electrical waveforms of these spikes all have a similar shape, so the acoustic variables are probably represented by the number of spikes and their arrival times. The arrival times of spikes represent physical variables which are interpreted in the brain. Ion-channel mechanisms underpin spike generation. Bats use intertwined acoustic and neural signal processing with various feedback control loops achieving an unparalleled performance: the animals are capable of satisfying all informational needs pertinent to their highly mobile, predatory lifestyles based on their (active and passive) sonar alone. Different system components have to work in an integrated fashion (e.g., shape of the emitting and receiving antennae, feature extraction by the neural code, adaptive mobility, and reshaping of the antennae) and require the interplay in a highly coupled system. The biosonar system of bats uses intertwined acoustic and neural signal processing to extract the features of interest required for the bat’s survival. . The acoustic processing is performed by the interaction between shape and position of the emitting and receiving antennae and the sound fields. The bats head uses panning and tilting of the neck and independent panning and tilting of each of the two ear structures. Biosonar systems found in nature (mostly in bats and dolphins) have long been recognized as exceptional examples of powerful yet parsimonious natural perception/sensorimotor systems.
Scientific investigations using biomimetics of various systems used in the bat sonar system which have been developed over the last 10 years have been successful in reproducing a few of several interesting aspects to biosonar function, none of them has managed to integrate all known functional features of a bat head so far. Being much larger than natural biosonar systems renders these models incapable of duplicating the diffraction effects which facilitate sonar sensing in their biological counterparts. . In any case, the generated spike codes have not come close to being a quantitatively correct description of what propagates in the auditory nerve of bats. Bats use for scanning their surroundings changing the instantaneous frequency of their chirps and hence in a way that is entirely independent of any moving parts.
Bat echolocation depends on transduction, where signals are mechanically amplified. 3
The Hodgkin–Huxley equation
Te Hodgkin–Huxley model explains several properties of the action potential that ensure faithful transmission of information from the cell body to axon terminals. First, action potentials are all or none. When a stimulus-induced neuronalmembrane depolarization is below the threshold, the action potential does not occur. When depolarization exceeds the threshold, the waveform of the actionpotential is determined by the timing of Na+ and K+ conductance changes and the relative concentrations of Sodium Na+ and Potassium K+ inside and outside the cell, which remain mostly constant for any given neuron. (Te Na+ infux and K+ efflux during an action potential cause very small changes in intracellular, and even smaller changes in extracellular, Na+ and K+ concentrations.) To a first approximation, action potentials assume the same form in response to any suprathreshold stimulus. Second, action potentials are regenerative—they propagate without attenuation in amplitude. Suppose that an action potential occurs at a particular site on the axon. Te rising phase creates a substantial membrane depolarization, which spreads down the axon and brings an adjacent region to threshold, which in turn does so for its adjacent downstream region, and so on
In this way, the action potential propagates in a similar form continuously and faithfully down the axon toward its terminals. Third, action potentials propagate unidirectionally in the axon, from the cell body to the axon terminals. When an action potential occurs at a given site on the axon (for example, site A in Figure above), in principle depolarization should also spread up the axon toward the cell body (site Z) in addition to spreading down the axon toward the axon terminals (site B). However, the delayed activation of the K+ channels and the inactivation of the Na+ channels combine to create a refractory period after an action potential has just occurred, during which time another action potential cannot be reinitiated. Because the action potential normally initiates at the axon initial segment and passes through Z before reaching A, another action potential cannot immediately back-propagate from A to Z. This refractory period ensures that the action potential normally propagates only from the cell body down the axon to its terminals, not in the reverse direction. In most projection neurons, whose axons form synapses on distant target neurons, the action potential first arises at the axon initial segment, where voltage-gated Na+ channel density per unit membrane area is the highest; this high channel density lowers the threshold for action potential initiation. The axon initial segment is a critical site for the integration of depolarizing and hyperpolarizing synaptic potentials from the dendrites and the cell body. After initiation, action potentials travel unidirectionally along the axon toward its terminals. At the initiation site, however, action potentials can in principle travel in both directions; indeed, in some mammalian neurons, action potentials can back-propagate to dendrites, which, like axons, contain voltage-gated Na+ and K+ channels. In artificial situations where experimenters electrically stimulate the axon or its terminals, action potentials can propagate in a retrograde direction from axon terminals to the cell body, producing so-called antidromic spikes, which can be recorded from the cell body. However, antidromic spikes have not been found to occur under physiological conditions in vivo.
Altogether, these three properties make the action potential an ideal means to transmit information faithfully from neuronal cell bodies across long distances to their axon terminals. But since action potentials are all or none, the size of action potentials cannot encode information about the stimulus. Rather, the information is usually encoded by the rate (number of action potentials per unit time) or the timing of action potentials in response to a stimulus. The spike rate is limited by the refractory period. Some neurons, such as fast-spiking inhibitory neurons in the mammalian cortex, can fire up to 1000 Hz, or one action potential per millisecond; the interval between these action potentials is shorter than the refractory period of many neurons. This requires specializing the ion channels so the action potential is repolarized quickly and the refractory period is complete in time for the next action potential. Tus, ion channel properties (such as Na+ channel inactivation and the delayed opening of K+ channels) have been selected during evolution to ensure unidirectional propagation of action potentials, and neurons with high spike rates use specialized ion channels with fast kinetics. The broad range of possible spike rates expands the information coding capacity of individual neurons.
There may be as many as 10,000 neuron cell types. Processing so much information requires a lot of neurons. How many? Well, "best estimates" indicate that there are around 200 billion neurons in the brain alone! And as each of these neurons is connected to between 5,000 and 200,000 other neurons, the number of ways that information flows among neurons in the brain is so large, it is greater than the number stars in the entire universe! While we are considering numbers, it is worth noting that there are as many as 50 times more glia than neurons in our CNS! Glia (or glial cells) are the cells that provide support to the neurons. In much the same way that the foundation, framework, walls, and roof of a house prove the structure through which run various electric, cable, and telephone lines, along with various pipes for water and waste, not only do glia provide the structural framework that allows networks of neurons to remain connected, they also attend to the brain's various house keeping functions (such as removing debris after neuronal death).
There are three kinds of neurons: motor neurons (for conveying motor information), sensory neurons (for conveying sensory information), and interneurons (which convey information between different types of neurons). The second and third parts are processes — structures that extend away from the cell body. Generally speaking, the function of a process is to be a conduit through which signals flow to or away from the cell body. Incoming signals from other neurons are (typically) received through its dendrites. The outgoing signal to other neurons flows along its axon. A neuron may have many thousands of dendrites, but it will have only one axon. The fourth distinct part of a neuron lies at the end of the axon, the axon terminals. These are the structures that contain neurotransmitters. Neurotransmitters are the chemical medium through which signals flow from one neuron to the next at chemical synapses.
To support the general function of the nervous system, neurons have unique capabilities for intracellular signaling (communication within the cell) and intercellular signaling (communication between cells). To achieve long distance, rapid communication, neurons have special abilities for sending electrical signals (action potentials) along axons. This mechanism, called conduction, is how the cell body of a neuron communicates with its own terminals via the axon. Communication between neurons is achieved at synapses by the process of neurotransmission.
Conduction
To begin conduction, an action potential is generated near the cell body portion of the axon. An action potential is an electrical signal very much like the electrical signals in electronic devices. But whereas an electrical signal in an electronic device occurs because electrons move along a wire, an electrical signal in a neuron occurs because ions move across the neuronal membrane. Ions are electrically charged particles. The protein membrane of a neuron acts as a barrier to ions. Ions move across the membrane through ion channels that open and close due to the presence of neurotransmitter. When the concentration of ions on the inside of the neuron changes, the electrical property of the membrane itself changes. Normally, the membrane potential of a neuron rests as -70 millivolts (and the membrane is said to be polarized). The influx and outflux of ions (through ion channels during neurotransmission) will make the inside of the target neuron more positive (hence, de-polarized). When this depolarization reaches a point of no return called a threshold, a large electrical signal is generated. This is the action potential.
This signal is then propagated along the axon (and not, say, back to its dendrites) until it reaches its axon terminals. An action potential travels along the axon quickly, moving at rates up to 150 meters (or roughly 500 feet) per second. Conduction ends at the axon terminals. Axon terminals are where neurotransmission begins. Hence, it is at axon terminals where the neuron sends its OUTPUT to other neurons. At electrical synapses, the OUTPUT will be the electrical signal itself. At chemical synapses, the OUTPUT will be neurotransmitter.
Neurotransmission
Neurotransmission (or synaptic transmission) is communication between neurons as accomplished by the movement of chemicals or electrical signals across a synapse. For any interneuron, its function is to receive INPUT "information" from other neurons through synapses, to process that information, then to send "information" as OUTPUT to other neurons through synapses. Consequently, an interneuron cannot fulfill its function if it is not connected to other neurons in a network. A network of neurons (or neural network) is merely a group of neurons through which information flows from one neuron to another. At electrical synapses, two neurons are physically connected to one another through gap junctions. Gap junctions permit changes in the electrical properties of one neuron to effect the other, and vice versa, so the two neurons essentially behave as one. Electrical neurotransmission is communication between two neurons at electrical synapses. 7
Action potentials are the result of currents that pass through ion channels in the cell membrane.
The velocity of action potentials is crucial for the right timing in information processing and depends on the dynamics of ion channels studding the axon, but also on its geometrical properties.
An experimenter as an external observer can evaluate and classify neuronal firing by a spike count measure – but is this really the code used by neurons in the brain? In other words, is a cortical neuron that receives signals from a sensory neuron only looking at and reacting to the number of spikes it receives in a time window of, say, 500 ms? A fly can react to new stimuli and change the direction of flight within 30-40 ms. This is not long enough for counting spikes and averaging over some long time window. The fly has to respond after a postsynaptic neuron has received one or two spikes. Humans can recognize visual scenes in just a few hundred milliseconds, even though recognition is believed to involve several processing steps. Again, this does not leave enough time to perform temporal averages on each level.
Biomimetics of Bat echolocation :
A New Bat-Like Sensor Captures Images Using Echolocation
https://www.allaboutcircuits.com/news/new-bat-like-sensor-captures-images-using-echolocation/
Team uses biomimicry of bats to help drones navigate in the dark, dust or smoke
https://techxplore.com/news/2018-06-team-biomimicry-drones-dark.html
1. https://www.nature.com/articles/416683a
2. https://www.thenewatlantis.com/publications/why-minds-are-not-like-computers
3. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.123.7084&rep=rep1&type=pdf
4. https://hackaday.com/tag/echolocation/
5. https://courses.cit.cornell.edu/bionb4240/Reprints/JonesTeeling2006_TREE.pdf
6. https://www.fcav.unesp.br/Home/departamentos/morfologia/ELISABETHCRISCUOLOURBINATI/materialdidatico/5---2018-auditory-mechanisms-of-echolocation-in-bats.pdf
7. http://www.mind.ilstu.edu/curriculum/neurons_intro/neurons_intro.php
https://en.wikipedia.org/wiki/Animal_echolocation
https://reasonandscience.catsboard.com/t3073-biological-computation-by-evolution-or-design
Artificial Intelligence performing algorithmic calculations are always the product of human engineering. It has never been demonstrated that a machine can become a mind. A computer can perform many different procedures; process algorithms — a set of distinct, well-defined steps. Selection sort accomplishes a task with one basic operation that, when performed over and over, completes the whole task. It also has a well-defined start state and end state, which can be referred to as its input and output. Algorithms have a well-defined set of steps for transforming input to output. So anyone who executes an algorithm will perform the same steps, and an algorithm’s output for a given input will be the same every time it is executed An algorithm doesn’t necessarily have to involve repetition, but any task performed on a large set of data usually will use such repeated steps, known as “loops.”
Take, for example, books that have titles composed of known characters, allowing for alphabetization; the shelf has an ordering (beginning to end, or left to right); the books are objects that can fit onto the shelf and be moved about; and so on. Only these few properties are relevant for the purposes of the algorithm. The sorter needs to know nothing about the full nature of a book in order to execute the algorithm — you need only have knowledge of shelf positions, titles, and how titles are ordered relative to one another. This abstraction is useful because the objects involved in the algorithm can easily be represented by symbols that describe only these relevant properties.
These two forms of abstraction are at the core of what enables the execution of procedures on a computer. At the level of its basic operations, a computer is both extremely fast and exceedingly stupid. The power of the computer derives not from its ability to perform complex operations, but from its ability to perform many simple operations very quickly. Any complex procedure that a computer performs must be reduced to the primitive operations that a computer can execute, which may require many levels at which the procedure is broken down into simpler and still simpler steps.
To do so, you must be able to represent the problem in terms that the computer can understand — but the computer only knows what numbers and memory slots are, not titles or shelves. The solution is to recognize that there is a correspondence between the objects that the computer understands and the relevant properties of the objects involved in the algorithm: for example, numbers and titles both have a definite order. You can use the concepts that the computer understands to symbolize the concepts of your problem: assign each letter to a number so that they will sort in the same way (1 for A, 26 for Z), and write a title as a list of letters represented by numbers; the shelf is in turn represented by a list of titles. You can then reduce the steps of your sorting job into steps at the level of simplicity of the computer’s basic operations. If you do this correctly, the computer can execute your algorithm by performing a series of arithmetical operations. (Of course, getting the computer to physically move your boss’s books is another matter, but it can give you a list ordered the way your boss wanted.)
This is why the computer is sometimes called a “symbol-manipulation machine”: what the computer does is manipulate symbols (numbers) according to instructions that we give it. The physical computer can thus solve problems in the limited sense that we imbue what it does with a meaning that represents our problem.
It is worth dwelling for a moment on the dualistic nature of this symbolism. Symbolic systems have two sides: the abstract concepts of the symbols themselves, and an instantiation of those symbols in a physical object. This dualism means that symbolic systems and their physical instantiations are separable in two important (and mirrored) ways. First, a physical object is independent of the symbols it represents: Any object that represents one set of symbols can also represent countless other symbols. A physical object and a symbolic system are only meaningfully related to each other through a particular encoding scheme. Thus it is only partially correct to say that a computer performs arithmetic calculations. As a physical object, the computer does no such thing — no more than a ball performs physics calculations when you drop it. It is only when we consider the computer through the symbolic system of arithmetic, and the way we have encoded it in the computer, that we can say it performs arithmetic.
A program is independent of the hardware that executes it; it could run just as well on many other pieces of hardware that work in very different ways. But a program is dependent on some physical representation in order to execute. 2
Bat echolocation, signal processing, bioengineering in its finest.
Echolocating animals produce high-frequency sounds through the mouth and use the arrival time, intensity, and frequency content of echo returns to determine the distance, direction, and features of objects in the environment.
By computing the time of reflection of modulates echoes, the bat is able to recognize the object and its distance. Over 1,000 species of bats echolocate with signals produced in their larynges. They use diverse sonar signal designs. Populations of neurons in the bat central nervous system respond selectively to the direction and delay of sonar echoes. In addition, premotor neurons in the bat brain are implicated in the production of sonar calls, along with the movement of the head and ears. Audio-motor circuits, within and across brain regions, lay the neural foundation for acoustic orientation by echolocation in bats. . The active component of echolocation is the animal’s production of sounds that reflect from objects in the environment, and perception is based on information processed from echo returns. The brain does a lot of processing and manipulation of incoming data to build a model of the outside environment.6
Bat's signal strength diminishes with the square of the distance, both going out and coming back. So the outgoing signals must be quite loud for the return echoes to be detectable. But there’s a problem. To pick up the weak return echoes, bat ears have to be extremely sensitive. But such sensitive ears would be wrecked by the loudness of the outgoing signals.
Bats have three distinct design features that must all work together at the same time their voice box, sensitive ears, and unique brains that allow them to understand the signals of echolocation.
So what to do? Bats amazingly turn off their ears during each outgoing chirp and turn them on again to catch each return echo. Ten to 200 times a second!
Question: In order for this system to work, had the mechanism of turning their ears off and on not to be fully developed right from the beginning?
Another problem: Typically there’s a zillion bats around, all creating these echos simultaneously. How can they distinguish their own from all those others? Well, they can, because each has its own distinctive signal. Their brain software masters this too, sorting their own echoes from all the background noise.
Question: How did they "learn" to generate an individual signal in order to distinguish their echoes from background noise? Trial and error?
Today, we all know that bats use echolocation to catch their prey. We take it for granted and see no reason why this fact should be awe-inspiring. Richard Dawkins in The Blind watchmaker page 43:
Donald Griffin tells a story of what happened when he and his colleague Robert Galambos first reported to an astonished conference of zoologists in 1940 their new discovery of the facts of bat echolocation. One distinguished scientist was so indignantly incredulous that he seized Galambos by the shoulders and shook him while complaining that we could not possibly mean such an outrageous suggestion. Radar and sonar were still highly classified developments in military technology, and the notion that bats might do anything even remotely analogous to the latest triumphs of electronic engineering struck most people as not only implausible but emotionally repugnant.
We can only understand it at a level of artificial instrumentation, and mathematical calculations on paper, we find it hard to imagine a little animal doing it in its head. Echo-sounding by bats is just one of the thousands of examples that I could have chosen to make the point about good design. Animals give the appearance of having been designed by a theoretically sophisticated and practically ingenious physicist or engineer, but there is no suggestion that the bats themselves know or understand the theory in the same sense as a physicist understands it. The bat should be thought of as analogous to the police radar trapping instrument, not to the person who designed that instrument. The designer of the police radar speed-meter understood the theory of the Doppler Effect, and expressed this understanding in mathematical equations, explicitly written out on paper. The designer's understanding is embodied in the design of the instrument, but the instrument itself does not understand how it works. The instrument contains electronic components, which are wired up so that they automatically compare two radar frequencies and convert the result into convenient units - miles per hour. The computation involved is complicated, but well within the powers of a small box of modem electronic components wired up in the proper way. Of course, a sophisticated conscious brain did the wiring up (or at least designed the wiring diagram), but no conscious brain is involved in the moment-to-moment working of the box.
Our experience of electronic technology prepares us to accept the idea that unconscious machinery can behave as if it understands complex mathematical ideas. This idea is directly transferable to the workings of living machinery. A bat is a machine, whose internal electronics are so wired up that its wing muscles cause it to home in on insects, as an unconscious guided-missile homes in on an airplane. So far our intuition, derived from technology, is correct. But our experience of technology also prepares us to see the mind of a conscious and purposeful designer in the genesis of sophisticated machinery. One form of the argument makes direct use of the extreme sense of wonder which we all feel when confronted with highly complicated machinery, like the detailed perfection of the echolocation equipment of bats.
Dawkins claim: It is this second intuition that is wrong in the case of living machinery. In the case of living machinery, the 'designer' is unconscious natural selection, the blind watchmaker. It isn't true that the whole perfect work must have been achieved simultaneously. It isn't true that each part is essential for the success of the whole. A simple, rudimentary, half-cocked eye/eat/ echolocation system/cuckoo parasitism system, etc., is better than none at all. How did echolocation get its start? Any animal that can hear at all may hear echoes. Blind humans frequently leam to make use of these echoes. A rudimentary version of such a skill in ancestral mammals would have provided ample raw material for natural selection to build upon, leading up by gradual degrees to the high perfection of bats. Where X is some organ too complex to have arisen by chance in a single step, then according to the theory of evolution by natural selection it must be the case that a fraction of an X is better than no X at all; One hundred and twenty five years on, we know a lot more about animals and plants than Darwin did, and still not a single case is known to me of a complex organ that could not have been formed by numerous successive slight modifications.
Remarkably, besides bats, dolphins, and whales, two birds, the oil-bird and the cave swiftlet also use echolocation, and these two genera have developed the same technology independently of each other. But furthermore, several different kinds of mammals, for instance, shrews, rats, tenrecs and seals seem to use echoes to a small extent, as blind humans do. Acoustic features such as Doppler shift compensation, whispering echolocation, and nasal emission of sound each show multiple convergent origins in bats. 5
Reply: From a computational neuroscience perspective, bats are remarkable because of the very short timescale on which they operate. The barrage of returning sonar echoes from a bat's near-environment lasts approximately 30 milliseconds following a sonar emission with the echo from a specific target lasting, at most, a few milliseconds.
From an engineering standpoint, biosonar systems (e.g. bats and dolphins) have inspired the design of very sophisticated sonar and radar systems that can map distant surfaces and track targets with great precision. Even with powerful mathematical tools and decades of experience, however, our best systems still do not rival the perceptual capabilities of dolphins. Many bats demonstrate incredible aerial agility, flying in complete darkness through branches and caves while hunting evasive insects. These animals perform such tasks in real-time with a total power consumption (including flight) measured in Watts, not hundreds of Watts. In addition to the ability to navigate in complete darkness by echolocation, both bats and dolphins live in very social environments using echolocation in group situations without any obvious problems with interference. All of these capabilities are highly desired by current military programs developing unmanned-aerial vehicles (UAV) especially since many of the target environments are in places where Global Positioning System (GPS) signals are unavailable and obstacle locations are not mapped.”
Bats use echolocation to see objects in front of them. They emit an ultrasonic pulse around 20 kHz (and up to 100 kHz) and then sense the pulses as they reflect off an object and back to the bat. 4
The echolocation process of bats involves three phases to search and capture prey: search phase, approach phase and terminal phase. During the search phase, the bat will start to hunt for prey by emitting the pulse at low rate with frequency around 10Hz. Then, the pulses have to get shorter as the time between the pulse and echo is decreased in order to avoid overlap when the bat spots and gets nearer to the specific prey during the approach phase. In this phase too, pulse emission rate gets steadily increased up to 200 per second since the bat keeps updating the position of the prey. In the terminal phase, the frequency of emitted pulses upsurges more than 200 Hz as the pulse emission rate also starts to accelerate at only a fraction of millisecond long just before the prey is netted
Many bats locate insect prey by emitting ultrasonic pulses and detecting the echoes. By detecting small Doppler shifts ( the change in frequency of a wave in relation to an observer who is moving relative to the wave source ) in the frequency of the returning pulses, the bat's nervous system can discern acoustic 'texture' and so distinguish prey from inanimate objects. Such elegant calculations are based on underlying computational mechanisms.
The Bat echolocation mechanism depends on electrical and mechanical engineering, robotics and control as well as neurological signal processing and sonar signal analysis.
Algorithms, finite sequences of well-defined, computer-implementable instructions, facilitate competent, or even near-optimal, computation; calculations of arithmetical steps that follow a well-defined model. These computations are carried out using biological molecules and cells. In bats, acoustic variables are represented in areas of the bat brain by the electrical 'spikes' emitted by neurons. The electrical waveforms of these spikes all have a similar shape, so the acoustic variables are probably represented by the number of spikes and their arrival times. The arrival times of spikes represent physical variables which are interpreted in the brain. Ion-channel mechanisms underpin spike generation. Bats use intertwined acoustic and neural signal processing with various feedback control loops achieving an unparalleled performance: the animals are capable of satisfying all informational needs pertinent to their highly mobile, predatory lifestyles based on their (active and passive) sonar alone. Different system components have to work in an integrated fashion (e.g., shape of the emitting and receiving antennae, feature extraction by the neural code, adaptive mobility, and reshaping of the antennae) and require the interplay in a highly coupled system. The biosonar system of bats uses intertwined acoustic and neural signal processing to extract the features of interest required for the bat’s survival. . The acoustic processing is performed by the interaction between shape and position of the emitting and receiving antennae and the sound fields. The bats head uses panning and tilting of the neck and independent panning and tilting of each of the two ear structures. Biosonar systems found in nature (mostly in bats and dolphins) have long been recognized as exceptional examples of powerful yet parsimonious natural perception/sensorimotor systems.
Scientific investigations using biomimetics of various systems used in the bat sonar system which have been developed over the last 10 years have been successful in reproducing a few of several interesting aspects to biosonar function, none of them has managed to integrate all known functional features of a bat head so far. Being much larger than natural biosonar systems renders these models incapable of duplicating the diffraction effects which facilitate sonar sensing in their biological counterparts. . In any case, the generated spike codes have not come close to being a quantitatively correct description of what propagates in the auditory nerve of bats. Bats use for scanning their surroundings changing the instantaneous frequency of their chirps and hence in a way that is entirely independent of any moving parts.
Bat echolocation depends on transduction, where signals are mechanically amplified. 3
The Hodgkin–Huxley equation
Te Hodgkin–Huxley model explains several properties of the action potential that ensure faithful transmission of information from the cell body to axon terminals. First, action potentials are all or none. When a stimulus-induced neuronalmembrane depolarization is below the threshold, the action potential does not occur. When depolarization exceeds the threshold, the waveform of the actionpotential is determined by the timing of Na+ and K+ conductance changes and the relative concentrations of Sodium Na+ and Potassium K+ inside and outside the cell, which remain mostly constant for any given neuron. (Te Na+ infux and K+ efflux during an action potential cause very small changes in intracellular, and even smaller changes in extracellular, Na+ and K+ concentrations.) To a first approximation, action potentials assume the same form in response to any suprathreshold stimulus. Second, action potentials are regenerative—they propagate without attenuation in amplitude. Suppose that an action potential occurs at a particular site on the axon. Te rising phase creates a substantial membrane depolarization, which spreads down the axon and brings an adjacent region to threshold, which in turn does so for its adjacent downstream region, and so on
In this way, the action potential propagates in a similar form continuously and faithfully down the axon toward its terminals. Third, action potentials propagate unidirectionally in the axon, from the cell body to the axon terminals. When an action potential occurs at a given site on the axon (for example, site A in Figure above), in principle depolarization should also spread up the axon toward the cell body (site Z) in addition to spreading down the axon toward the axon terminals (site B). However, the delayed activation of the K+ channels and the inactivation of the Na+ channels combine to create a refractory period after an action potential has just occurred, during which time another action potential cannot be reinitiated. Because the action potential normally initiates at the axon initial segment and passes through Z before reaching A, another action potential cannot immediately back-propagate from A to Z. This refractory period ensures that the action potential normally propagates only from the cell body down the axon to its terminals, not in the reverse direction. In most projection neurons, whose axons form synapses on distant target neurons, the action potential first arises at the axon initial segment, where voltage-gated Na+ channel density per unit membrane area is the highest; this high channel density lowers the threshold for action potential initiation. The axon initial segment is a critical site for the integration of depolarizing and hyperpolarizing synaptic potentials from the dendrites and the cell body. After initiation, action potentials travel unidirectionally along the axon toward its terminals. At the initiation site, however, action potentials can in principle travel in both directions; indeed, in some mammalian neurons, action potentials can back-propagate to dendrites, which, like axons, contain voltage-gated Na+ and K+ channels. In artificial situations where experimenters electrically stimulate the axon or its terminals, action potentials can propagate in a retrograde direction from axon terminals to the cell body, producing so-called antidromic spikes, which can be recorded from the cell body. However, antidromic spikes have not been found to occur under physiological conditions in vivo.
Altogether, these three properties make the action potential an ideal means to transmit information faithfully from neuronal cell bodies across long distances to their axon terminals. But since action potentials are all or none, the size of action potentials cannot encode information about the stimulus. Rather, the information is usually encoded by the rate (number of action potentials per unit time) or the timing of action potentials in response to a stimulus. The spike rate is limited by the refractory period. Some neurons, such as fast-spiking inhibitory neurons in the mammalian cortex, can fire up to 1000 Hz, or one action potential per millisecond; the interval between these action potentials is shorter than the refractory period of many neurons. This requires specializing the ion channels so the action potential is repolarized quickly and the refractory period is complete in time for the next action potential. Tus, ion channel properties (such as Na+ channel inactivation and the delayed opening of K+ channels) have been selected during evolution to ensure unidirectional propagation of action potentials, and neurons with high spike rates use specialized ion channels with fast kinetics. The broad range of possible spike rates expands the information coding capacity of individual neurons.
There may be as many as 10,000 neuron cell types. Processing so much information requires a lot of neurons. How many? Well, "best estimates" indicate that there are around 200 billion neurons in the brain alone! And as each of these neurons is connected to between 5,000 and 200,000 other neurons, the number of ways that information flows among neurons in the brain is so large, it is greater than the number stars in the entire universe! While we are considering numbers, it is worth noting that there are as many as 50 times more glia than neurons in our CNS! Glia (or glial cells) are the cells that provide support to the neurons. In much the same way that the foundation, framework, walls, and roof of a house prove the structure through which run various electric, cable, and telephone lines, along with various pipes for water and waste, not only do glia provide the structural framework that allows networks of neurons to remain connected, they also attend to the brain's various house keeping functions (such as removing debris after neuronal death).
There are three kinds of neurons: motor neurons (for conveying motor information), sensory neurons (for conveying sensory information), and interneurons (which convey information between different types of neurons). The second and third parts are processes — structures that extend away from the cell body. Generally speaking, the function of a process is to be a conduit through which signals flow to or away from the cell body. Incoming signals from other neurons are (typically) received through its dendrites. The outgoing signal to other neurons flows along its axon. A neuron may have many thousands of dendrites, but it will have only one axon. The fourth distinct part of a neuron lies at the end of the axon, the axon terminals. These are the structures that contain neurotransmitters. Neurotransmitters are the chemical medium through which signals flow from one neuron to the next at chemical synapses.
To support the general function of the nervous system, neurons have unique capabilities for intracellular signaling (communication within the cell) and intercellular signaling (communication between cells). To achieve long distance, rapid communication, neurons have special abilities for sending electrical signals (action potentials) along axons. This mechanism, called conduction, is how the cell body of a neuron communicates with its own terminals via the axon. Communication between neurons is achieved at synapses by the process of neurotransmission.
Conduction
To begin conduction, an action potential is generated near the cell body portion of the axon. An action potential is an electrical signal very much like the electrical signals in electronic devices. But whereas an electrical signal in an electronic device occurs because electrons move along a wire, an electrical signal in a neuron occurs because ions move across the neuronal membrane. Ions are electrically charged particles. The protein membrane of a neuron acts as a barrier to ions. Ions move across the membrane through ion channels that open and close due to the presence of neurotransmitter. When the concentration of ions on the inside of the neuron changes, the electrical property of the membrane itself changes. Normally, the membrane potential of a neuron rests as -70 millivolts (and the membrane is said to be polarized). The influx and outflux of ions (through ion channels during neurotransmission) will make the inside of the target neuron more positive (hence, de-polarized). When this depolarization reaches a point of no return called a threshold, a large electrical signal is generated. This is the action potential.
This signal is then propagated along the axon (and not, say, back to its dendrites) until it reaches its axon terminals. An action potential travels along the axon quickly, moving at rates up to 150 meters (or roughly 500 feet) per second. Conduction ends at the axon terminals. Axon terminals are where neurotransmission begins. Hence, it is at axon terminals where the neuron sends its OUTPUT to other neurons. At electrical synapses, the OUTPUT will be the electrical signal itself. At chemical synapses, the OUTPUT will be neurotransmitter.
Neurotransmission
Neurotransmission (or synaptic transmission) is communication between neurons as accomplished by the movement of chemicals or electrical signals across a synapse. For any interneuron, its function is to receive INPUT "information" from other neurons through synapses, to process that information, then to send "information" as OUTPUT to other neurons through synapses. Consequently, an interneuron cannot fulfill its function if it is not connected to other neurons in a network. A network of neurons (or neural network) is merely a group of neurons through which information flows from one neuron to another. At electrical synapses, two neurons are physically connected to one another through gap junctions. Gap junctions permit changes in the electrical properties of one neuron to effect the other, and vice versa, so the two neurons essentially behave as one. Electrical neurotransmission is communication between two neurons at electrical synapses. 7
Action potentials are the result of currents that pass through ion channels in the cell membrane.
The velocity of action potentials is crucial for the right timing in information processing and depends on the dynamics of ion channels studding the axon, but also on its geometrical properties.
An experimenter as an external observer can evaluate and classify neuronal firing by a spike count measure – but is this really the code used by neurons in the brain? In other words, is a cortical neuron that receives signals from a sensory neuron only looking at and reacting to the number of spikes it receives in a time window of, say, 500 ms? A fly can react to new stimuli and change the direction of flight within 30-40 ms. This is not long enough for counting spikes and averaging over some long time window. The fly has to respond after a postsynaptic neuron has received one or two spikes. Humans can recognize visual scenes in just a few hundred milliseconds, even though recognition is believed to involve several processing steps. Again, this does not leave enough time to perform temporal averages on each level.
Biomimetics of Bat echolocation :
A New Bat-Like Sensor Captures Images Using Echolocation
https://www.allaboutcircuits.com/news/new-bat-like-sensor-captures-images-using-echolocation/
Team uses biomimicry of bats to help drones navigate in the dark, dust or smoke
https://techxplore.com/news/2018-06-team-biomimicry-drones-dark.html
1. https://www.nature.com/articles/416683a
2. https://www.thenewatlantis.com/publications/why-minds-are-not-like-computers
3. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.123.7084&rep=rep1&type=pdf
4. https://hackaday.com/tag/echolocation/
5. https://courses.cit.cornell.edu/bionb4240/Reprints/JonesTeeling2006_TREE.pdf
6. https://www.fcav.unesp.br/Home/departamentos/morfologia/ELISABETHCRISCUOLOURBINATI/materialdidatico/5---2018-auditory-mechanisms-of-echolocation-in-bats.pdf
7. http://www.mind.ilstu.edu/curriculum/neurons_intro/neurons_intro.php
https://en.wikipedia.org/wiki/Animal_echolocation