cs 344 ai boredom talk muhahaha

8/9/2019 CS 344 AI Boredom Talk Muhahaha

1/25

Manas Joglekar (07005006)Harsh Pareek (07005007)

Viraj Deshpande (07005008)Mehul Jain (07005021)Rohit Singh (07005034)

Manas Joglekar (07

005006)Harsh Pareek (07005007)

Viraj Deshpande (07005008)Mehul Jain (07005021)Rohit Singh (07005034)


2/25

Outline

y Introductiony Definition

y

Psychology, Philosophy and Biologicaly Does intelligence require boredom?

y Boredom as awareness of self

y Modeling Boredom

y Behavioral testsy Using AI

yApplications

y Future Directions

y Introductiony Definition

y

Psychology, Philosophy an

d Biologicaly Does intelligence require boredom?

y Boredom as awareness of self

y Modeling Boredom

y Behavioral testsy Using AI

yApplications

y Future Directions


3/25

Definition

yBoredom is an emotional state experienced during periods lackingactivity or when individuals are uninterested in the opportunitiessurrounding them.

yThe first record of the word boredom is in the novel Bleak House byCharles Dickens, written in 1852

How people have described boredom:

y Boredom [is] the desire for desires

yLeo Tolstoy, Anna Karenina

yA man can stand almost anything except a succession of ordinarydays.

yGoethe, the great 18th century German poet, dramatist, and scientist,described boredom as the mother of invention.

yMan is the only animal that can be bored.yErich Fromm, The Sane Society (1955)

yBoredom is an emotional state exp

erienced during periods lackingactivity or when individuals are uninterested in the opportunitiessurrounding them.

yThe first record of the word boredom is in the novel Bleak House byCharles Dickens, written in 1852

How ople e escribedboredom:

y Boredom [is] the desire for desires

yLeo Tolstoy, Anna arenina

yAman can stand almost anything except a succession of ordinarydays.

yGoethe, the great 18th century German poet, dramatist, and scientist,described boredom as the mother of invention.

yMan is the only animal that can be bored.yErich Fromm, The ane ociety (1955)


4/25

Psychology

yBoredom is an unpleasant, transient affective state in which theindividual feels a pervasive lack of interest and difficultyconcentrating on the current activity.

yPositively, a response to a moderate challenge for which thesubject has more than enough skill

y

Boredom has been traditionally associated with clinicaldepression and attention disorders.

yBoredom is an unpleasant, trans

ient affective state in which theindividual feels a pervasive lack of interest and difficultyconcentrating on the current activity.

yPositively, a response to a moderate challenge for which thesubject has more than enough skill

y

Boredom has been traditionally associated with clinicaldepression and attention disorders.


5/25

Mental State as a Function of Skill and

Challenge Levels- Mihaly Csikszentmihalyi

Mental State as a Fun

ction of Skill and

Challenge Levels- Mihaly Csikszentmihalyi


6/25

Philosophysituational boredom

chronic boredom- existential boredom

Arthur Schopenhauer used the existence of boredom in anattempt to prove the vanity of human existence,

"...for if life, in the desire for which our essence and existenceconsists, possessed in itselfa positive value and real content, therewould be no such thing as boredom: mere existence wouldfulfilland satisf yus .

situational oredom

chronic oredom- existential oredom

Arthur Schopenhauer used the ex

istence of boredom in anattempt to prove the vanity of human existence,

"...for if life, in the desire for which our essence and existenceconsists, possessed in itselfa positive value and real content, therewould be no such thing as boredom: mere existence would fulfilland satisfy us .


7/25

Default Network

y The bored brain : incredibly active, generates daydreamsand engages in mental time travel.

y An elaborate electrical conversation between the frontand rear parts of the mind.

y In humans: generates spontaneous thoughts duringmind-wandering, believed to be an essential componentofcreativity.

Defaul etwork

y The bored brain : incredibly active, generates daydreamsand engages in mental time travel.

y An elaborate electrical conversation between the frontand rear parts of the mind.

y n humans: generates spontaneous thoughts duringmind-wandering, believed to be an essential componentofcreativity.


8/25

Boredom and limits of AI

Can we build machines which can think?

Can we build machines which can think like humanbeings?

Can computers imitate humanbehaviour(intelligence)?

Can we identify an aspect of intelligence whichseems to be really difficult for computers to achieve?

Boredom and limits of AI

Can we build machines which can think?

Can we build machines which can think like humanbeings?

Can computers imitate humanbehaviour(intelligence)?

Can we identify an aspect of intelligence whichseems to be really difficult for computers to achieve?


9/25

Can computers be creative?

Creativity: ability to produce unanalyzable leaps ofimagination (creative sparks).

Is everyone creative or is it that only a few of us are?

"noncreative intelligence- a flat-out contradictionin terms.

Still vague

Can computers be creative?

Creativity: ability to produce unanalyzable leaps ofimagination (creative sparks).

Is everyone creative or is it that only a few of us are?

"noncreative intelligence- a flat-out contradictionin terms.

Still vague


10/25

Creativity as a network ofconcepts

having creativity is an automatic consequence of havingthe proper representation of concepts in the mind.

concepts derive all their power from their connectivity to

one another

Vaguely: concepts are the rules, facts and data available toan intelligent being.

What kind of objects have such a representation ofconcepts and what kinds do not?

Creativity as a network ofconcepts

having creativity is an automatic consequence of havingthe proper representation of concepts in the mind.

concepts derive all their power from their connectivity to

one another

Vaguely: concepts are the rules, facts and data available toan intelligent being.

What kind of objects have such a representation ofconcepts and what kinds do not?


11/25

Boring Multimedia

Computer Bug

Playing Cat

Sphex

Bori Multi edia

Computer Bug

Playing Cat

Sphex


12/25

SphexishnessHofstadter{Metamagical Themas Questing for the Essence of

Mind and Pattern(1982)}; Christopher von Bulow(2003)--Cant analyze its own behavior Never gets bored!

What is the mantis lacking?

a general sensitivity to patterns, an ability to spot patterns ofunanticipated types in unanticipated places at unanticipatedtimes in unanticipated media.(theres a pattern up there too did you recognize it?)

Human Beings: loop detectorlosing the disposition responsible for the loop.

Hofstadter{Metamagical Themas Questing for the Essence of

Mind and Pattern(1982)}; Christopher von Bulow(2003)--Cant analyze its own behavior Never gets bored!

What is the mantis lacking?

a general sensitivity to patterns, an ability to spot patterns ofunanticipated types in unanticipated places at unanticipatedtimes in unanticipated media.(theres a pattern up there too did you recognize it?)

Human Beings: loop detectorlosing the disposition responsible for the loop.


13/25

Can we create antisphexish machines?

Does your computer get bored?

Requires self-watching at various levels of abstraction.

Should a computer yawn after every similar set of instructions itexecutes?

Watching one's own internal microscopic patterns is bound to beboring.

The computer needs to identify regular behavior due to hugepatterns of activity at a higher level

Can we create antisphexish machines?

Does your computer get bored?

Requires self-watching at various levels of abstraction.

Should a computer yawn after every similar set of instructions itexecutes?

Watching one's own internal microscopic patterns is bound to beboring.

The computer needs to identify regular behavior due to hugepatterns of activity at a higher level


14/25

Contd

Self-watching : monitoring changes in data structures ,filteringand recording interesting aspects in otherdatastructures(watchers).

Shouldnt we set up a second level of data structures to monitorthese watchers?

-Territory for infinite regress!

Thats not all! Shouldnt we have a watcher for this infinite chain?

Self-watching : monitoring changes in data structures ,filteringand recording interesting aspects in otherdatastructures(watchers).

Shouldnt we set up a second level of data structures to monitorthese watchers?

-Territory for infinite regress!

Thats not all! Shouldnt we have a watcher for this infinite chain?


15/25

"Minds,Machines, and Gdel -J.R. Lucas

Machines cannot think like humans!

Humans cannot get into an infinite loop but a machine will!

Knowledge Chain: A conscious being knows something, knowsthat he knows it, and that he knows that he knows that he knowsit, and so on.

Argument against AI: A machine can be made in a manner of

speaking to consider its performance, but it cannot take thisinto account without thereby becoming a different machine,namely the old machine with a new part added.

"Minds,Machines, and Gdel -J.R. Lucas

Machines cannot think like humans!

Humans cannot get into an infinite loop but a machine will!

Knowledge Chain: A conscious being knows something, knowsthat he knows it, and that he knows that he knows that he knowsit, and so on.

Argument against AI: A machine can be made in a manner of

speaking to consider its performance, but it cannot take thisinto account without thereby becoming a different machine,namely the old machine with a new part added.


16/25

Conclusion: perfectly antisphexish

machines are impossible

Remember the Halting Problem?Is there any program inspecting other programs before they runpredicting termination?

Closely related (diagonalization)

Difference between the two problems(AI survives):

a self-watching program(for AI): not really so concerned with the

mathematical perfection but with likelihood of survival in acomplex world;

It seems that self watching is essential for an AI program (e.g.learning) but perfect self watching is something one cannot

achieve. We have to settle with some approximation that works.

Conclusion: perfectly antisphexish

machines are impossible

Remember the Halting Problem?Is there any program inspecting other programs before they runpredicting termination?

Closely related (diagonalization)

Difference between the two problems(AI survives):

a self-watching program(for AI): not really so concerned with the

mathematical perfection but with likelihood of survival in acomplex world;

It seems that self watching is essential for an AI program (e.g.learning) but perfect self watching is something one cannot

achieve. We have to settle with some approximation that works.


17/25

Turing Test

yA human judge engages in a natural language conversationwith one human and one machine, each of which tries to

appear human. All participants are placed in isolated locationsyIf the judge cannot reliably tell the machine from the human,the machine is said to have passed the test.

yA human judge engages in a natural language conversationwith one human and one machine, each of which tries to

appear human. All participants are placed in isolated locationsyIf the judge cannot reliably tell the machine from the human,the machine is said to have passed the test.


18/25

Measuring Boredomy Non-verbal tests: Audio-visual feedback techniques

y Facial expression

y Sitting positiony Pressure map on chair

y Head height

y Leaning / slouching

y Resistance across

points on skiny Heart beat

y Blood pressure

y Audio feedback

Measuri Boredomy Non-verbal tests:Audio-visual feedback techniques

y Facial expression

y Sitting positiony Pressure map on chair

y Head height

y Leaning / slouching

y Resistance across

points on skiny Heart beat

y Blood pressure

y Audio feedback


19/25

Processing audio-visual feedbacky Computational techniques used:

y Hidden Markov models

y Feed-forward neural networks

y Support vector machines

y Fuzzy ARTMAP network

Processi audio-visual feedbacky Computational techniques used:

y Hidden Markov models

y Feed-forward neural networks

y Support vector machines

y FuzzyARTMAP network


20/25

Measuring Boredom in Humans

Boredom Proneness ScaleCognitive scale proposed in 1986 byFarmer, R. & Sundberg, N. D.

This has been widely studied to obtain variation of proneness to boredomacross gender, age, race, work, personality, etc.

Relevance of this scale has been highly studied and various modificationshave been suggested.Instead of one-dimensional bipolar scale: 5 dimensions can be extractedfrom the survey:

y External Stimulation, Internal Stimulation, Constraints, Affective

Response, and Perception of Time

Other proposed scales:yEASI-III Temperament Surveyy36-point Academic Boredom scale

Measuri Boredom i umans

Boredom Proneness ScaleCognitive scale proposed in 1986 byFarmer, R. & Sundberg, N. D.

This has been widely studied to obtain variation of proneness to boredomacross gender, age, race, work, personality, etc.

Relevance of this scale has been highly studied and various modificationshave been suggested.nstead of one-dimensional bipolar scale: 5 dimensions can be extracted

from the survey:

y External Stimulation, nternal Stimulation, Constraints, Affective

Response, and Perception of Time

Other proposed scales:yEAS - Temperament Surveyy36-pointAcademic Boredom scale


21/25

Some results


22/25

Non-rigid scales:

Difficult to design and heavy computation.yModeling boredom and flow using web activity ofusers:

yType of activity:y

Information retrievalyCreating webpagesyCommunication: Social/personal (email/chatting)yPlaying games

yResponse time/ immediate feedbacky

Design of websites most visitedyClear goals

Non-ri idscales:

Difficult to design and heavy computation.yModeling boredom and f low using web activity ofusers:

yType of activity:y

nformation retrievalyCreating webpagesyCommunication: Social/personal (email/chatting)yPlaying games

yResponse time/ immediate feedbacky

Design of websites most visitedyClear goals


23/25

ApplicationsyAcademic boredom:

y Intelligent tutoring systemsy Systems with adaptive strategies based on audio-visual

feedback from students

y Predicting dropouts from e-coursesy Time invariant vs. Time variant parameters

yWork environment boredom:y Detecting boredom before significant drop in

performance is observed

y Self-organizing task allocation among agentsy 2 agent experiment

yAcademic boredom:y Intelligent tutoring systems

y Systems with adaptive strategies based on audio-visualfeedback from students

y Predicting dropouts from e-coursesy Time invariant vs. Time variant parameters

yWork environment boredom:y Detecting boredom before significant drop in

performance is observed

y Self-organizing task allocation among agentsy 2 agent experiment


24/25

Future Directions

Should a system stop a computation if it notices a pattern?

Or perhaps use a pattern to speed up computation?

Eg. Learning rate in typical AI systems

Stop when not making progress

Future Directions

Should a system stop a computation if it notices a pattern?

Or perhaps use a pattern to speed up computation?

Eg. Learning rate in typical AI systems

Stop when not making progress


25/25

References

http://en.wikipedia.org/wiki/Boredom http://www.bartneck.de/publications/2008/emotionAndAI/index.html

http://chaos.uncc.edu/~heckel/papers/JacobsEtAlHRI2009.pdf Metamagical Themas: Questing For The Essence Of Mind And

Pattern by: Douglas Hofstadter http://uwf.edu/svodanov/boredom/bps.htm - A Boredom

Proneness test on a 7-point scale Farmer, R. & Sundberg, N. D. (1986). Boredom proneness: The

development and correlates of a new scale.Journal ofPersonalityAssessment, 50, 417

http://scienceblogs.com/cortex/2009/03/boredom.php http://mightymouse.brynmawr.edu/~dblank/papers/epirob-

abs.pdf

http://en.wikipedia.org/wiki/Boredom http://www.bartneck.de/publications/2008/emotionAndAI/index.html

http://chaos.uncc.edu/~heckel/papers/ acobsEtAlHRI2009.pdf Metamagical Themas: uesting For The Essence OfMindAnd

Pattern by: Douglas Hofstadter http://uwf.edu/svodanov/boredom/bps.htm - ABoredom

Proneness test on a 7-point scale Farmer, R. & Sundberg, N. D. (1986). Boredom proneness: The

development and correlates of a new scale.J

ournalofersonalityAssessment, , 417 http://scienceblogs.com/cortex/2009/03/boredom.php http://mightymouse.brynmawr.edu/~dblank/papers/epirob-

abs.pdf

cs 344 ai boredom talk muhahaha

Documents