Cloned from: M366 OU block 3

by bugsy


keywords:
Bookmark and Share



Front Back
  natural intelligence
  a broader, more inclusive view of intelligence, one that looks beyond the human sphere, and which we call natural intelligence
    anthropomorphism
    seeing intelligence only in behaviour that resembles our own, dismissing everything else as ‘instinct’ or some such word.
    goal-directed
    aimed towards some specific goal
    chemotaxis
Chemotaxis is the phenomenon in which somatic cells, bacteria, and other single-cell or multicellular organisms direct their movements according to certain chemicals in their environment
    taxis
response by an organism to a directional stimulus or gradient of stimulus intensity.
top-down organisation.
A single directing intelligence sits at the top of a hierarchy of tasks and control.
    bottomup
no overall control, no managing brain, no design, no obvious breakdown of tasks.
    Drives .
    At the root of every form of animal behaviour there seem to be basic drives. Living creatures have purposes in what they do. We are all familiar with the complexity of human motives.
 Recognition
 with increasing sophistication comes an increasing ability to recognise and discriminate.
    Classification
comes the related ability to classify or categorise the things that are sensed
Response
reactions  may appear straightforward at first sight, they conceal deep complexity
    Communication .
    Many animals do have elementary ‘languages’ systems of signalling to one another. But without syntax, subtle semantics or verbal association, these have nothing remotely resembling the power of our languages.
    Learning
    Learning seems crucial for intelligence. An animal that is inflexible, that has no capacity to change in response to experience, will be a poor contender in the struggle for existence
    carbon chauvinism
    There are no grounds for asserting that biology must depend on organic molecules:
    Representation (explicit representation of knowledge in some symbolic form) and search.
    two main principles, discussed in Block 2, that guide the construction of conventional artificial intelligence programs.
    knowledge elicitation
    Human experts doctors, chemists, geologists are quizzed in lengthy sessions of knowledge elicitation to draw out their knowledge
    formalism
    knowledge represented in rule-like formalism which can then be transferred to a computer where it can be processed by algorithms.
  Propositional knowledge
  knowledge that can be expressed explicitly in the form of propositions, such as ‘The Battle of Hastings was in 1066’, or ‘The light from distant galaxies shows a red shift.’ It is sometimes called ‘ knowing that’, since you can precede each proposition with the phrase ‘I know that ...
  Non-propositional knowledge
  knowledge that can’t be expressed in this way, usually because it is manifested in some form of ability or skill. Examples include ‘I know how to swim’, ‘I know how to speak French’, etc. So non-propositional knowledge is often called ‘knowing how’.
    biologically inspired computing , or often nouvelle AI.
    The project to bring insights about the mechanisms underlying natural intelligence to difficult computational problems
    pheromones
    special chemicals
    recruitment
    deposits are attractive to other ants of the colony, and an individual ant will be impelled to move towards one if it senses it nearby
    self-organisation
    Since no single agent is in control of the organised activity of the system, we refer to this evolution of systematic behaviour as self-organisation.
    SENSE PLANACT cycle.
    data is fed from the sensors to a planning system which manipulates this virtual map, and then sends action commands to its motors.
    Parsimony
    There are no complex internal representations, and almost certainly no symbols of any kind
    Fluent coupling .
    the insect’s sensory apparatus, the muscles which move its body, and the environment all working together as a closely coupled system.
    stigmergy
    One insect modifies the environment in a certain way and another then senses and responds to this change
    neurons
    Specialised neurons make up the sensory system of every animal, while other specialised neurons activate the muscles.
    cellular automata .
    a large class of computer systems
  blocks
  stable structures
  blinkers
  oscillating structures
  glider
  patterns that move across the grid, reconstituting themselves after a certain fixed number of steps
  glider guns
  structures which regularly generate other structures, such as glider guns, which periodically emit a new glider;
  eaters
  configurations that destroy other patterns such as gliders that collide with them.
emergence
 We can state that structures (gliders, blocks, puffer-trains, beehives, etc.) of The Game of Life emerge from interactions (the rules of birth and death) between simple components (individual cells).

    interacting components, being in some way more than the sum of these parts.
bifurcations
  abrupt changes of behaviour
heteropathic laws heteropathic effects

 homeopathic effect is simply the sum of effects of each of the parts of the system acting individually.     However, Mill argued that heteropathic effects cannot be understood in this way: the outcome is more than just the sum of the effects of the parts, and can only be explained in terms of certain supplementary heteropathic laws, which he described as ‘laws of combined agency [which] are not compounded of the laws of the separate agencies’.
    reductive
they account for phenomena wholly in terms of simpler things
    strong and weak emergence
    Strongly emergent phenomena simply cannot be predicted from knowledge of their low-level base: they are something radically new and different. By contrast, a high-level phenomenon is weakly emergent from a low-level domain when facts about that phenomenon are unexpected, but could in principle be deduced from facts about the low-level domain, given enough knowledge, time and reasoning power.
  non linear
a nonlinear system is any problem where the variable(s) to be solved for cannot be written as a linear combination of independent components
    adaptation
    as the world changes, we observe that the living things that inhabit it change too.
    learning
    one of the most significant types of adaptation that will occur in any animal, and especially so in humans.
 behaviourism

    conditioned
animals can be trained to develop a reflex response, such as salivation, to a certain event, such as a bell ringing
operant conditioning,
    responses that have some effect in the surrounding world (pressing a pedal, for example) are reinforced by a reward, such as a food pellet
  precocial species
  fully developed capabilities at birth
  altricial species
  The young of these creatures are more or less helpless at birth, and may take a long time to acquire the capabilities of an adult.
    evolutionary computation
    a form of computing inspired by the key ideas of Darwinian natural selection
    Scientific investigation .
    Building a computer model of a biological system (or any other system found in nature) helps us to understand the properties of natural systems. A computer model enables scientists to:  make predictions;  check to what extent the behaviour of the system corresponds with the natural reality;  vary key parameters of the model and so clarify their influence on the way the system works.
    Software design .
    it is possible to build computer systems, based on the four mechanisms of natural intelligence, that can solve complex practical problems in commerce, design, economics and medicine
 variables  rules
    a scientific or computer model isolates a few relevant variables, the values of which are related to one another according to known rules.
x of y cards Next >|