perturbance unifying research on emotion intrusive
play

Perturbance: Unifying research on emotion, intrusive mentation and - PowerPoint PPT Presentation

Perturbance: Unifying research on emotion, intrusive mentation and other psychological phenomena with AI Luc P. Beaudoin (Simon Fraser University. Conflict of Interest Declaration: CogZest, CogSci Apps Corp.) Sylwia Hyniewska (University of


  1. Perturbance: Unifying research on emotion, intrusive mentation and other psychological phenomena with AI Luc P. Beaudoin (Simon Fraser University. Conflict of Interest Declaration: CogZest, CogSci Apps Corp.) Sylwia Hyniewska (University of Bath & Bioimaging Research Center Warsaw) http://hyniewska.jeb.be Eva Hudlicka (Psychometrix Associates, University of Massachusetts, Amherst) Paper presented at the Symposium on Computational Modelling of Emotion: Theory and Application at AISB–2017. Paper and Slides Available from http://summit.sfu.ca/item/16776. Presentation originally given on 2017-04-20 at University of Bath, and at University of Birmingham, England on 2017-04-24.

  2. Phenomena/Lit.’s to Unify • Some “Emo(ons” • “Intrusive thinking”, “Obsession” / OCD • “Rumina(on”, “worry” • Pain • addic(ons • Insomnia , depression, anxiety, mania, etc. • Memory: Effects of tes(ng ‘memory’ • Some mental effects of experiencing art Perturbance Beaudoin , Hyniewska, Hudlicka 2

  3. (One of) Sloman’s (Beautiful!) Theses • These types of emo(on will emerge as a bi- product of mo+ve processing mechanisms in autonomous agents. * Key concepts: • Mo+ve processing (mechanisms)** – ASen(on (mechanisms) – Emergence – Perturbance Beaudoin , Hyniewska, Hudlicka 3

  4. Perturbance Beaudoin , Hyniewska, Hudlicka 4

  5. Perturbance Beaudoin , Hyniewska, Hudlicka 5

  6. Perturbance Beaudoin , Hyniewska, Hudlicka 6

  7. Perturbance Beaudoin , Hyniewska, Hudlicka 7

  8. Emotion Producing Circumstances • Losing your full purse 😲 • Your parents’ pension having been wiped out (corp. bankruptcy due to exec. fraud) 😲 / 😪 / 😢 • Falling in love 😋 • & Deciding to leave your spouse & children 😠 • Finding you’ve got cancer • … Perturbance Beaudoin , Hyniewska, Hudlicka 8

  9. Pierre Curie Mind “ Does she ( Marie Sk ł odowska) love me? ” ❤ Perturbance Beaudoin , Hyniewska, Hudlicka 9

  10. “But these are not emotions” • 1992: Enter “Perturbance” Perturbance Beaudoin , Hyniewska, Hudlicka 10

  11. 1995-2008 (example progress) • Architecture-based ‘emo(on’ concept dis(nc(ons – Primary ‘emo(ons’ (Alarms) – Secondary ‘emo(on’ – ‘Ter(ary Emo(on’ = Perturbance . [This talk] • loss of control of execu%ve func%ons , not of a-en%on . • Several Ph.D theses • Work on aSachment • So]ware Libraries (SIMAGENT, etc.) Perturbance Beaudoin , Hyniewska, Hudlicka 11

  12. 2010+ • Perturbance (concept, relabelled as ‘ter(ary emo(on’) virtually ignored • Why is the Perturbance/ H-CogAff beauty asleep? Perturbance Beaudoin , Hyniewska, Hudlicka 12

  13. Focusing Phenomenon “Whatever we put our attention on will grow stronger in our life” Maharishi Mahesh Yogi • “Where your aSen(on goes, energy flows” • “Focusing Illusion” Kahneman et al. • “Anchoring and Adjustment” • “Familiarity increases liking” (“Mere exposure effect”) Perturbance Beaudoin , Hyniewska, Hudlicka 13

  14. ❤ • ? • ?? • ??? Perturbance Beaudoin , Hyniewska, Hudlicka 14

  15. BENVOLIO: Be ruled by me. Forget to think of her . ROMEO: O, teach me how I should forget to think! Perturbance Beaudoin , Hyniewska, Hudlicka 15

  16. “this book offers insight into love, infatuation , madness, and all flavors of emotion in between.” Perturbance Beaudoin , Hyniewska, Hudlicka 16

  17. Unlike “Emotion” Yay! :) Unanimous! - Ignored :( BUT - Frijda yes, - No C of perturbance Perturbance Beaudoin , Hyniewska, Hudlicka 17

  18. Focus on Limerence Because … • ↑Clearly perturbant (aSen(on, mo(ve processing, ↑ insistence, intensity ) • ↑ Clear evolu(onary significance/func(ons • ↑ Clear commitment phenomenon ( Aubé, 200x; ACachment: Dean PeCers ). • Intermediate/LT Consciousness (Donald, 2001) – O]en lasts months, some(mes years • ↑ ‘Intense’ and ↑ Memorable (though infrequent) • → Happiness/misery (meaning, purpose, etc.) • ↑ Limerent (Poten(al) Par(cipants on Campus Perturbance Beaudoin , Hyniewska, Hudlicka 18

  19. How Can I Tell You? Wherever I am girl, I'm always walking with you I'm always walking with you But I look and you're not there Whoever I'm with, I'm always walking with you I'm always walking with you , And I'm sad that you can't hear Perturbance Beaudoin , Hyniewska, Hudlicka 19

  20. “Repetitive Thought” (‘RT’) • “Intrusive thinking” • “Obsession” / OCD • “Rumina(on”, “worry”; • Transdiagnos(c: – insomnia , – depression (“depressive rumina(on”, etc), – anxiety (“anxious rumina(on” , etc.), – mania, etc. • Pain, addic(ons, etc. (earlier slide) Perturbance Beaudoin , Hyniewska, Hudlicka 20

  21. Watkins (2008) “Constructive and unconstructive repetitive thought” Perturbance Beaudoin , Hyniewska, Hudlicka 21

  22. Watkins’ RT Concept’s Supposed Advantages • atheore(cal • clearer • non-evalua(ve • more inclusive • ↑ correlated w measures of worry&rumina(on Perturbance Beaudoin , Hyniewska, Hudlicka 22

  23. Watkins’ RT Concept’s Actual Disadvantages • atheore(cal: Problem! Cf. physics, chem . • clearer: Atheore+cal → ↑Ambiguity . • non-evalua(ve: What about normal IP ? • more inclusive: (a) Not; (b) undifferen+ated . • ↑ correlated w measures of worry&rumina(on; begs ques+on (psychology’s obsession) . Perturbance Beaudoin , Hyniewska, Hudlicka 23

  24. Big Software Development Projects Involve System Architects Who • Study sys. requirements • Design the en(re architecture • Consult with/to Components Developers • Do not implement mechanisms themselves • “Architecture dominates mechanism” Perturbance Beaudoin , Hyniewska, Hudlicka 24

  25. Future Research • Today 99% (??) of research papers about [all forms of] RT are empirical! • Remedies/Goals – Equal amount of publica(on/effort from designer stance. – 80% (?) of empirical work: Use Arch.-based theory • Empiricists: use architecture-consultants • Beware of Math envy Perturbance Beaudoin , Hyniewska, Hudlicka 25

  26. Future Research (Driven) • AI: Engage psychologists successfully : – Empirical research – ASen(on, 1 consciousness, 5 mo(va(on 2 – Applica(ons (“Affec(ve redirec(on”, insomnia, emo(on) 3 . • Problem of mo(vator iden(ty * • Problem of mo(vator development (cf. J Mitchell’s talk). [tree] 4 * • How is insistence managed? Perturbance Beaudoin , Hyniewska, Hudlicka 26

  27. Why Perturbance (C) • Requires the designer stance • Promising research programme • Specified in terms of whole minds • The genera(ve power of mo(ve processing constructs • Framework to explain emo(on episodes • Promise of emergence (parsimony) Perturbance Beaudoin , Hyniewska, Hudlicka 27

  28. CogAff Project History (Part1) • A.S. 1978: Computer Revolu(on in Philosophy • A.S. 1981: Why robots will have emo(ons* • A.S. 1981: You don't need a so] skin to have a warm heart: Towards a computa(onal analysis of Mo+ves and Emo+ons* • A.S. (1987) Mo+ves, mechanisms and emo+ons • AS. (1991/3) Prolegomena to a Theory of * Communica(on and Affect 1994; 1997 Perturbance Beaudoin , Hyniewska, Hudlicka 28

  29. Related Work • Simon (2017-50). Mo+va+onal & emo(onal*… • Mandler on plans • Oatley & Johnson-Laird • Kruglanski (2016) and Baumeister (2016) papers on importance of mo(va(on for various problems of psychology (including emo(on) • Pessoa (2014, Cogni(ve Emo(onal Brain) Perturbance Beaudoin , Hyniewska, Hudlicka 29

Recommend


More recommend