Sam's blog

Day 025 - The trouble with qualia

Submitted by Sam on 15 June, 2011 - 01:25

Our brains have evolved to consciously parse only the most relevant and important information from our senses, making them more like filters than windows to the world, as Robert Ornstein observed in his 1986 book, Multimind. Through this filtering process comes subjectivity, and its attendant philosophical problems, notably qualia, or the ineffably subjective quality of conscious sensation; the perception of the sensation of the colour red, 'the redness of red'. An example of one of my own most distinctly identifiable qualia is my sensation of gut-wrenching dread whenever I hear a specific alarm-clock tone. Other people will have their own widely different feelings and experiences associated with this particular tone, and hence their own quale associated with it.

Qualia are deceptively complex because they inhabit the liminal space between the physical and the mental worlds, and are thus a central component to the mind-body problem. A great deal of philosophical debate has arisen from qualia, broadly centred around whether these subjective experiences can be described by purely physical information or not – i.e. whether they can ever exist without being experienced. Can I ever know what it is like to experience your sensation of the colour red? Whilst we have already heard a neuroscientist's answer to this question from Henry Markram at The Blue Brain Project, it is interesting to contrast his physicalist perspective with the numerous philosophical thought experiments which have been devised to answer this question, which I shall be looking at over the next few days.

In the meantime, here is a quick overview of qualia from Dr Ramachandran.

Like all components of living organisms, the evolution of conscious systems has been shaped by selection pressures which favour energy-efficient and resource-saving architectures. Conscious systems have been optimized by many generations of natural selection to process and store only the most relevant and essential information.

The human mind is an extremely compact conscious system, saving both space and time by perceiving and remembering a simplified but very organized version of the environment rather than a resource-hungry one-to-one objective replica. To defer to Stan Franklin once more, a useful analogy for visualizing this optimization is to compare a stored table of the first x prime numbers with a formula which calculates them on demand. The formula is much more compact, potentially requiring much less memory (as in bits) than a large look-up table, and so is preferable in systems where memory is at a premium. This compression results in a species-wide subjective experience of the real world, and one manifestation of it can be tested through psychophysical experiments.

Psychophysics is a branch of science concerned with the quantitative relationships between mental states and physical events and processes, exploring the intersection of the subjective with the objective in order to make subjective internal transformations measurable.

The Weber-Fechner law is an historically important psychophysical attempt to describe the relationship between the magnitudes of stimuli and their perceived intensity. By recording when a person was able to distinguish between two nearly identical stimuli, such as two similar weights, it was found that the smallest noticeable difference between them was approximately proportional to the intensity of the stimulus, so that the threshold for distinction varies according to the intensity. This means that a weight increase of a few grams on a 1kg weight is undetected, and is only perceived when a critically increased threshold mass is reached; and if the mass is doubled then the threshold is also doubled. This describes a logarithmic relationship between stimulus and perception, illustrating how the world our energy-efficient brains create is affected by internal transformations.

The Stevens' power law offers several refinements to the logarithmic relationship described by the Weber-Fechner law, showing that whilst our visual perception of length may be linear, our perception of brightness is logarithmic and our perception of pain is exponential, further revealing that what we perceive does not infallibly correlate with the physical world, but is instead mediated by the wiring of our brains.

Day 023 - Degrees of mind

Submitted by Sam on 13 June, 2011 - 00:27

By exhibiting over the course of its life a descending order of 'mind', moving from a small but adequate nervous system – a very low degree of mind – to almost no nervous system at all – a negligible degree of mind, the sea squirt's extraordinary metamorphosis draws attention to the possible existence of degrees of mind, a spectrum of complexity which would place bacteria at one extreme and humans at another. The following syllogism from Stan Franklin helps to clarify this particular definition of mind, which has a much wider reach than the dictionary definition provided by the OED, where the mind is described as a phenomenon peculiar only to human beings 1. Franklin sees mind as that which controls behaviour, noting that bacteria behave, and so must therefore possess some degree of mind in order to control their behaviour.

The sea squirt suggests that even very basic minds are expensive to maintain, requiring valuable energy and resources that are conserved at the earliest opportunity. Possessing a mind must therefore outweigh the costs of fuelling it, and so selective evolution must ensure that a balance is maintained between the degree of mind and the resources required to run it. As the complexity of mind increases, any energy-saving optimizations are preserved. The consequences of this architectural trait are important for an understanding of the human mind, and I'll look at some of them in tomorrow's blog.

  • 1. The OED definition of mind as a mental or physical faculty is: 'The seat of awareness, thought, volition, feeling, and memory; cognitive and emotional phenomena and powers considered as constituting a presiding influence; the mental faculty of a human being (esp. as regarded as being separate from the physical); (occas.) this whole system as constituting a person's character or individuality'. "mind, n.1". OED Online. March 2011. Oxford University Press. 12 June 2011 http://oed.com/view/Entry/118732?rskey=s5NKU5&result=1&isAdvanced=false

Day 022 - Sea squirts – no more deciding

Submitted by Sam on 11 June, 2011 - 21:47

The life cycle of the sea squirt features a remarkable metamorphic transition where the animal absorbs its own rudimentary brain. This extreme change is brought about by the sea squirt achieving the uniquely rare state of having no more decisions left to make.

Sea squirts are living representations of some of our most primitive ancestors, belonging to the same phylum that we do (Chordata) because of the basic vertebrate features which characterize them during their early life as larvae. In their larval form, sea squirts resemble small tadpoles, and are able to swim around freely. In this form, these small marine animals possess a notochord (a flexible rod-shaped “backbone”), a tail, a primitive eye, and a balance organ, making them an effective dispersal mechanism for finding a suitable habitat for their adult life. They swim around in this form until they find a suitably hard surface, perhaps a rock, a ship's hull or the back of a large crab, to which they will attach themselves by standing on their head and using specialized organs of attachment, papillae, to permanently adhere themselves.

Once attached to the solid substrate where they will spend the rest of their lives, the metamorphoses begins. Within a single day, the sea squirt absorbs most of its tail, its nerve cord and its notochord – all of the features that classify them as chordates. As their larval features quickly degenerate, the larval organs are reabsorbed and replaced with an adult set more suited to a sedentary life, and the sea squirt becomes “an entirely new animal” 1.

Firmly fixed in place, filter-feeding food particles from the water, the complex machinery of a brain is simply superfluous to the sea squirt's now sessile life. Having achieved its ultimate purpose of deciding once and for all what to do next, the brain is no longer needed, and so is absorbed. The sea squirt's metamorphosis can be seen as a stark validation of Stan Franklin's interpretation of the single purpose of mind being to choose what to do next.

  • 1. Dethier, Vincent G. "The Magic of Metamorphosis: Nature's Own Sleight of Hand." Smithsonian 17.2 (1986): 122

Day 021 - The what-to-do-next machine

Submitted by Sam on 10 June, 2011 - 23:47

So what is the purpose of all of this emergent behaviour? Is there any point to the intelligence that all of these independently limited components join together to achieve? For Stan Franklin, Interdisciplinary Research Fellow at The University of Memphis and co-director of the Institute for Intelligent Systems, all of the richness of mind is directed towards one easily expressible goal; to decide what to do next 1 .

On the most basic level, every flash of electro-chemical activity in a neural network is in essence a calculation, moving the system from one state to another, processing inputs into outputs. Because of this basic architecture, brains are constantly performing operations, locked into a cycle of deciding what to do next, as the neural network processes calculations over time to produce the conscious sensation of weighing-up facts to make a high-level decision. From a behavioural perspective, all of these operations are performed in order to effect the optimal next move for the organism's well-being, processing all available variables to produce a consensus of action which will, ideally, benefit the organism in some way.

To make informed decisions, thinking organisms must be hard-wired with some fundamental inherited values against which to measure all other decisions; without such a benchmark, one action is as good as another. At some basic level, all decision-making organisms must have some invariant autonomous goals, preferences and standards, against which all other decisions can be compared. Some identifiable goals might include: eat, drink, mate, avoid pain. Again, in order to satisfy these basic drives, the organism needs to be constructed either with some minimal knowledge necessary to achieve these goals, or with the means to learn how to do so.

  • 1. Franklin, Stan. Artificial Minds. Cambridge, MA: MIT, 1997

Day 020 - Superorganisms

Submitted by Sam on 9 June, 2011 - 21:44

An interesting analogue can be drawn between the emergent complexity of the local interaction of neurons and the collaborative behaviour of superorganisms such as ants. Collectively, both ants and neurons exhibit emergent group intelligence in a strikingly similar manner. Colonies of ants and neural networks achieve things that no individual component can achieve on its own, as neurons link together to produce circuits of logic gates with universal computational capabilities, and as ants collaborate to build and maintain highly structured societies and intricate nests of excavated tunnels and chambers.

Ants are well-known examples of eusocial animals, each performing highly-specialized tasks. Harvester ants, for example, can be categorized into four distinct categories, based on the tasks that they elect to perform: patrolling, foraging, waste management and nest maintenance. Based on local information, ants decide to switch between tasks, producing coordinated emergent behaviour and patterns of global work which create a very stable colony.

With this specialization and exceptionally well-organized division of labour, each ant in a colony can be seen as an individual component of a much larger system in much the same way that, for instance, a neocortical column can be seen as a discrete functional unit of the brain. In this way, ants and neural networks exhibit a strikingly similar form of distributed intelligence, whereby cooperative agents of minimal individual intelligence achieve system-level goals far beyond their individual capabilities. Through such compartmentalization and modularity, both brains and ant colonies have evolved into systems which are self-evidently highly robust, reliable and scalable.

The similarities between the two systems are extensive. Like neurons (and Reynolds' boids), ants act in accordance with simple rules and local information – they are not controlled by a centralized body which oversees the colony 1, but are instead reactive only to a small set of simple chemical signals released by other ants. At this level of abstraction, the ants communicate just as neurons communicate at synapses, where chemical neurotransmitter is released in a highly localized environment.

Neuro-structural reassembly, the brain's process of reducing the overall number of connections to leave only the most efficient, shares many common features with the emergent optimization of ant food-gathering. An ant collecting food leaves behind it a particular pattern of pheromones which other drones will smell and follow towards the food source, and over time the most direct path to the food will become the most successful as more and more ants follow it and reinforce it, forming their distinctive trail. This self-organizing, short-cut-taking behaviour bears a strong resemblance to how the brain wires itself during development.

Similarly, if an ant colony becomes overcrowded or damaged, scouts will leave the nest to search for a new site. Once more using pheromone trails, the ants 'decide' on a new site when a critical threshold of scouts 'agree' on a particular location for a new nest. Research 2 has shown a very similar process occurring in a monkey's visual cortex when it performed a visual discrimination task. Neuron activity gradually increases until a threshold is reached, and the monkey makes a decision.

Ant colonies provide a high-level metaphor for the organization and logic of the brain, usefully illustrating how comparatively simple individual elements can join together in a distributed, decentralized fashion to produce complex, intelligent behaviour.

  • 1. The queen ant is a deceptive term – she has no overall control over her children in the colony, and serves only to continually reproduce.
  • 2. Valeo, Tom. "Researchers Begin to Decode Decision-making Processes - Dana Foundation." Brain and Brain Research Information - Dana Foundation. Web. 09 June 2011. http://www.dana.org/news/brainwork/detail.aspx?id=14382.

Day 019 - Emergence

Submitted by Sam on 8 June, 2011 - 23:35

The induced procedural behaviour of optogenetically engineered fruit flies and the invariantly instinctive behaviour of the Sphex wasp are examples of how complex behaviour can be encoded programmatically, with fixed stimuli producing fixed responses. These fixed pathways can be seen as discrete 'programs' in the brain, a subset of the overall neural network devoted to a particular function. Through the interaction of many such discrete components, extremely complex behaviour can emerge; though the components themselves are not intelligent, their combined interaction can produce intelligent behaviour. This is the theory of emergence, and it is the animating force behind attempts to create artificially intelligent neural networks.

The classic example of high-level artificially emergent behaviour is Reynolds' Boids, a computer model developed in 1986 to simulate coordinated animal motion. Through the interaction of three basic rules, each simulated boid is steered into a flock. The simple rules which produce this behaviour are:

  • Separation: steer to avoid crowding local flockmates
  • Alignment: steer towards the average heading of local flockmates
  • Cohesion: steer to move toward the average position (centre of mass) of local flockmates

Unexpected behaviours such as flocks dividing and then reuniting to avoid obstacles emerge from the interaction of these rules, showing once more how simple instructions can produce organized behaviour.

Day 018 - Natural programmatic behaviour

Submitted by Sam on 8 June, 2011 - 00:05

Gero Miesenböeck produced a programmatic behavioural response in fruit flies by selectively re-wiring their neural circuitry, establishing an evidential link between the electrical states of individual neurons and specific behaviour. His findings offer micro-scale support to our understanding of instinctive behaviour, in particular the abundance of examples of fixed action paths that can be found in nature.

Fixed action patterns are hard-wired, invariant sequences which run until completion in response to a sensory stimulus, known as a sign stimulus or releaser. Whilst very similar to reflex actions, which are clearly instinctive, fixed action paths can be processed by the brain. Reflex actions, conversely, do not go through the brain, but instead trace a reflex arc which involves no processing by the neural networks of the brain at all.

A favourite example of an easily triggered fixed action pattern is found in some species of Sphex digger wasp. Adult females lay their eggs in a burrow, leaving them to hatch on their own. Before leaving them, the mother wasp captures and paralyzes a cricket, dragging it to the edge of the burrow. She then goes in alone, apparently checking for intruders before emerging to drag the cricket inside, leaving it close to her eggs to provide food for them when they hatch. This extremely complex sequence of behaviour, whilst seeming to show signs of planning and forethought, is actually a completely automated routine, a fixed action path which gives the illusion of intelligence. If an experimenter moves the cricket very slightly away from the burrow whilst she is inside, she will re-emerge, drag the cricket back to the threshold, and re-enter the burrow once more. This cycle of behaviour has been induced dozens of times, illustrating that the cricket is merely following a strict procedure in response to an initial stimulus, in this case placing the cricket in front of the burrow.

When one of Miesenböeck's graduate students, Susana Lima, engineered a fruit fly so that just two of its brain cells expressed a light-activated pore causing the fly to invariably take off in response to a flash of light, she had clearly found the cells which were responsible for initiating one of the fly's fixed action patterns. She was able to initiate the reflex using an internal neuronal stimulus, rather than the external releaser that is so readily identifiable in the Sphex wasp example.

Day 017 - Optogenetic control of the brain

Submitted by Sam on 7 June, 2011 - 01:14

So far we have seen projects committed to making computers work like brains, encoding the immense complexity of the biological world into the procedural logic of the digital. Waynflete Professor of Physiology at Oxford Gero Miesenböeck has approached the brain from the reverse perspective, developing genetic strategies to control the brain itself like a computer.

Miesenböeck is the founder of the emerging field of optogenetics, which exploits genetic engineering to create targeted cells in living animals which can be switched from one state to another by a flash of light. Optogenetic remote control of the brain is the subject of his TED talk, below.

Miesenböeck formulates the familiar mantra of projects like The Human Brain Project as "if we could record the activity of our neurons, we would understand the brain", but modulates the recipe through his own work as the more practical “if we could control the activity of some neurons, we would learn much about the brain”. Rather than taking apart the brain and rebuilding it to see how it works, Miesenböeck's research focuses on controlling the brain in order to understand it.

In order to control the electrical impulses of the brain, which in turn control both behaviour (as we have known since Galvani's experiments on frogs in the 18th century) and thought, Miesenböeck re-engineered selected neural elements to become responsive to light, a non-invasive, diffuse signal. Through his optogenetic engineering he turned selected nerve cells into receivers that allowed him to control their function through a flash of light onto the brain. Using this technique, Miesenböeck successfully simulated an unpleasant memory in a fruit fly, causing it to 'remember' to walk away from a certain odour every time it encountered it. Through repeated experimentation with differentially activated nerve cells, Miesenböeck was able to narrow this behaviour down to twelve specific neurons in the fly's brain, which he identifies as the brain's 'critic', defining policy for the fly's 'actors', such as the circuits which control leg movement.

Miesenböeck's optogenetical control of behaviour provides arresting experimental evidence for the physical, mechanistic understanding of the mind that connectionist models are attempting to simulate in silicon.

Day 016 - Neuroinformatics

Submitted by Sam on 6 June, 2011 - 01:10

Projects like The Blue Brain Project and The HBP require a consolidated and consistent body of neuroscientific data against which to calibrate their virtual brain models, and so standardizing and aggregating the findings from international, interdisciplinary research is consequently vital to the success of computational neuroscience. The International Neuroinformatics Coordinating Facility, or INCF, was established in 2005 in response to this growing need, aiming to develop and steward a highly structured international informatics network of standardized neuroscientific databases and models. Neuroinformatics is therefore a discipline which encompasses all levels and scales of neuroscience (from genes to behaviour), facilitating the sharing and accessibility of data, tools and techniques amongst the neuroscientific community.

The task of collecting and integrating the field's data has become increasingly difficult as it has gained in popularity and diversity, with many thousands of publications representing many different countries, disciplines and methodologies released every year. The principle goal of the INCF is to systematize this immense wealth of material into databases which can be efficiently analyzed and modelled by researchers worldwide. The INCF also promotes standards and common frameworks in order to address infrastructural issues within the international neuroscientific community itself, which arise from the inherent difficulties of different labs working with multiple experimental techniques. By developing programmes for data-sharing, data-reuse, modelling, reporting and recording, the INCF benefits not only neuroscience as a whole, but also specifically supports efforts to build digital simulations of the brain by providing an addressable body of knowledge against which to test and verify virtualized behaviour.

Syndicate content
Drupal theme by Kiwi Themes.