3 How intelligent are modern chatter bots Do they meet Turings prediction - PowerPoint PPT Presentation

1 / 14
About This Presentation
Title:

3 How intelligent are modern chatter bots Do they meet Turings prediction

Description:

Chatter bots. Are pieces of software. ... No chatter bots can claim to have human capacities of reasoning and deduction, ... on which the chatter bots is based ... – PowerPoint PPT presentation

Number of Views:74
Avg rating:3.0/5.0
Slides: 15
Provided by: csh96
Category:

less

Transcript and Presenter's Notes

Title: 3 How intelligent are modern chatter bots Do they meet Turings prediction


1
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • Chatter bots
  • Are pieces of software.
  • Using natural language processing, their
    program attempts to identify what a user says -
    the input - then applies a variety of different
    methods, from pattern matching and keyword
    identification to neural networks and fuzzy
    logic, to produce an appropriate answer.
  • No chatter bots can claim to have human
    capacities of reasoning and deduction, and by
    no means are yet able to reproduce all the
    complexity of the human brain.

2
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • How intelligent are chatterbots depend
  • on
  • the technology on which the chatter bots is based
    on
  • the quality of the knowledge base - the brain -
    that has been developed by the botmaster
  • what you expect from them ?

3
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • Developers have found a number of solutions which
  • enable chatter bots to imitate human
    conversational
  • ability, based upon a range of tricks for
    avoiding
  • Difficult questions and sticking to areas that
    they
  • can handle, which in some ways is a proof of
  • intelligence in itself.

4
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • These include
  • making controversial statements to provoke a
    response
  • agreeing with the user, rather than being
    non-committal
  • repeating user input in their answers to make it
    seem as if they are following the conversation
  • remembering and reusing past topics of
    conversation
  • changing the topic when they don't understand

5
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • Turing Predication
  • Turing predicted that machines would be capable
    of
  • being as human as humans by the end of the
    twentieth
  • century... but it seems as if it will be some
    years yet
  • before this dream becomes reality.

6
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • Example A.L.I.C.E Chatterbots
  • A.L.I.C.E Statement
  • are not related to what I was chatting.
  • Sometime say I was not designed for.
  • Not remembering the past conversations.

7
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • Conclusion
  • More and more effort is going into chatterbots
    research
  • and development.
  • Although acquiring human intelligence remains a
    distant
  • dream, chatterbots technology can only keep
    improving,
  • to meet the ever-growing need for communication
    and
  • information.

8
3) How intelligent are modern chatter bots? Do
they meet Turings prediction?
  • References
  • www.agentland.com
  • www.alicebot.org
  • www.stetson.edu

9
5) What would or do you regard as evidence of
intelligence in a machine?
  • Evidence of intelligence in a machine can be
  • easily found if we can answer the question
  • below
  • What do you consider intelligence to be?

10
5) What would or do you regard as evidence of
intelligence in a machine?
  • Approach
  • A human cannot be intelligent in all the areas.
  • E.g.
  • An engineer is knowledgeable in his own area
  • and if we ask any question related to medical
  • field he might not be able to answer. But, as an
  • engineer he will be able direct us to some
  • medical representative who will be able to
  • answer.

11
5) What would or do you regard as evidence of
intelligence in a machine?
  • Evidence of intelligence in a machine
  • Pass the Turing test.
  • Ability to
  • understand and learn
  • solve problem
  • make own decisions.
  • adapt

12
5) What would or do you regard as evidence of
intelligence in a machine?
  • Execute efficient search mechanisms to carry out
    complex computations
  • Should be able to exhibit proper responses to
    various stimuli
  • Should have some kind of memory and be able to
    interact learn from new environments

13
5) What would or do you regard as evidence of
intelligence in a machine?
  • They should be able to act on their own
  • Machines have been built to emulate human
    behaviour
  • Current chatter bots have shown a good sense of
    humor

14
5) What would or do you regard as evidence of
intelligence in a machine?
  • E.g. Asimo
  • (Advanced Step in Innovative Mobility)
  • Look like humans and have similar intelligence
    and abilities of human.
  • Capable of performing most common household
    chores.
  • In addition, the androids will function as
    round-the-clock security guards capable of
    phoning the police or fire departments in case of
    emergency.
Write a Comment
User Comments (0)
About PowerShow.com