Intelligence as an Emergent Phenomenon - PowerPoint PPT Presentation

1 / 21
About This Presentation
Title:

Intelligence as an Emergent Phenomenon

Description:

Bodily Information 'Evolution embodies information in every part of every organism' ... Distributed storage of physical properties ... – PowerPoint PPT presentation

Number of Views:22
Avg rating:3.0/5.0
Slides: 22
Provided by: johf
Category:

less

Transcript and Presenter's Notes

Title: Intelligence as an Emergent Phenomenon


1
Intelligence as an Emergent Phenomenon
  • J Duncan

2
Daniel Dennetts Kinds of Minds
  • Intelligence, from the lowest forms to the
    highest is an emergent phenomena
  • Connectionism and distributed information storage
    is the means by which this is realized
  • Humans are separated only from other animals by
    use of language and our ability to hypothesize
    about identity

3
Origins
  • Earliest replicators - Macromolecules
  • Perform random walks for needs
  • Use chemical lock and key mechanisms
  • Reasons but no reason formulators, or reason
    representers, or even reason appreciators

4
The Need-to-Know Principle
  • Enforced by the cost of design in nature
  • The cheapest, least intensively designed system
    will be discovered first by Mother Nature, and
    myopically selected.
  • Emergent behavior is much cheaper than directed,
    efficient design

5
Problems and Benefits
  • Problems
  • Not necessarily the best or smallest design
  • Legacy code
  • Unexpected benefits
  • Recycling code
  • Random mutations of unused material

6
Plants and Bias
  • Respond phenotypically to stimuli on evolutionary
    timescale
  • timescale chauvinism / spatial bias - size
  • What is the actual role of speed in the
    phenomenon of the mind?
  • Mind barrier to be crossed?

7
Mind Speed Requirements?
  • Relative speed
  • Minimum for usefulness due to environment
  • Within an ecosystem, fast wins over slow
  • Emergent properties might also manifest at some
    speed but not below

8
Sentience vs. Sensitivity
  • mere intentional systems vs. genuine minds
  • Sentience lowest grade of consciousness
  • Some missing factor?
  • Looking for x

9
Processing
  • Plants distributed processing
  • Animals and plants operations distinct of
    central processing
  • Cellular level - defenses
  • Biochemical -gt electric impulse nervous system
  • Separation of function from composition

10
Functionalism
  • Multiple realizations
  • Artificial minds should be constructible if we
    can isolate what minds do
  • Process information
  • Disagreement over media

11
Evolution of Transduction
  • Fly by wire
  • Efficient information transmission vs. force
    transmission
  • Feedback helps
  • Time lags
  • Isolate the control system mind?

12
Double Tansduction
  • Myth of Double Tansduction percepts -gt
    Information (neural signals) -gt thought
    (Decartes)
  • since peripheral activity was mere sensitivity,
    there had to be some more central place where the
    sentience was created.
  • Artificial Intelligence

13
Connectionism
  • The idea that the network itself- by virtue of
    its intricate structure could assume the role of
    the inner Boss and thus harbor consciousness,
    seems preposterous. Initially
  • Information processing Information storage
    Physical structure of the network

14
Bodily Information
  • Evolution embodies information in every part of
    every organism
  • Physical information - world
  • Not necessarily found in the brain
  • Distributed storage of physical properties
  • This only takes you so far short range
    intentionality

15
Microagents
  • Evolution from proximal systems to distal
    systems
  • Organisms composed of microagents, organized
    into dozens or hundreds or thousands of
    subsystems. Each of these tiny agents can be
    conceived of as an utterly minimal intentional
    system
  • Nodes in a connectionist network cells and
    groups of cells connected in networks

16
Tower of Generate-and-Test
  • Darwinian Creatures (field testing)
  • Skinnerian Creatures (reinforcement learning,
    NNs) simple invertebrates?
  • Popperian Creatures (preselection of actions,
    inner model)
  • Gregorian Creatures (tool users and tool
    refiners)

17
Design pitfalls
  • Top-down and bottom-up design in Cognitive
    Science
  • Working up we tend to be unimpressed
  • Working down we subtract particularly human
    features of our experience as inessential.
  • We may tend to overlookthe possibility that we
    are subtracting, on one path, the very thing we
    are seeking on the other.

18
No Distinction?
  • A conservative hypothesis Sentience comes in
    every imaginable grade or intensity, from the
    simplest and most robotic, to the most
    exquisitely sensitive, hyper-reactive human.
  • Tool use
  • Among the preeminent tools, Gregory reminds us,
    are what he calls mind tools words

19
As-if Decision Making
  • Systems of perception non-Boolean circuits
  • Braitenbergs vehicles
  • By installing dozens or hundreds or thousands of
    such circuits in a single organism, elaborate
    life-protecting activities can be reliably
    controlled, all without anything happening inside
    the organism that looks like thinking specific
    thoughts. There is plenty of as if decision
    making

20
What is unique about humans?
  • The capacity we have to frame, and even under
    most circumstances test, hypotheses about
    identity
  • it takes a Gregorian creature who has language
    among its mind tools. But in order to use
    language, we have to be specially equipped with
    the talents that permit us to extract these mind
    tools from the social environment in which they
    reside.

21
Discussion
Write a Comment
User Comments (0)
About PowerShow.com