Colloquium organized by the Institute of Psychoanalysis and Management
and by the University Paris I Panthéon Sorbonne
on the subject
THE BIAS OF HUMAN INTELLIGENCE
AND ARTIFICIAL INTELLIGENCE
Paris, 25 et 26 novembre 2021
IAE de Paris-Sorbonne Business School
Call for PapersSubmission of your choice in English or French. For other languages (Spanish, German, Italian... etc.), submissions must be made in English or in French Call_NDL_for_Papers_IPM_IAE_2021_appendix_English_1.pdf From decision biases to algorithmic biases. The contributions of the psychoanalytical perspective in the study of cognitive, perceptual, emotional, individual and collective biases perceived in the context of digital organizations. Informing decision making and innovation processes. Digital transformation is now the motor of innovation within organizations. It nevertheless contributes to the diffusion of decision-making and cognitive biases. These biases impact decision making and innovation processes within digital organizations. The aim of the conference is to gather contributions on this theme, providing a perspective supported by the contributions of psychoanalysis. It aims to bring together papers exploring the problem of algorithmic biases in Psychoanalysis & Management. These biases are still little known. On purpose, one can mention biases linked to the programming of algorithms, economic biases, biases aiming to orient behaviors. The multiplication of algorithmic biasesThe functioning of organizations relies more and more on automated processes and digital systems based on Artificial Intelligence, which use algorithms, especially "learners". However, these are often perceived as being subject to various "biases", which can be detrimental to the neutrality, fairness and equity of treatment. These biases occur "when the heuristics and data in algorithms are influenced by the values of the humans involved in their collection, selection and/or use" (Nissenbaum, 2001). They can thus influence the results of search engines, learning data processing, social network services, individual or organizational profiling, etc. These biases are the subject of various classifications (Caliskan, Bryson and Narayanan, 2017, Lambrecht and Tucker 2017, d'Ignazio and Klein, 2020), which generally distinguish between "programmer bias" (anchoring bias) and "bias in the algorithm, When they distort predictive justice or predictive finance deep learning applications, these biases can lead to serious dysfunctions in organizations and/or behaviors. The contributions of behavioral finance (FC)Recent studies on algorithmic biases build on older work on cognitive, perceptual, emotional and group biases. Since the 1970s, FC has been observing psychological biases in the decision making of investors and financial intermediaries. Their decisions are subject to substantial, limited and/or procedural rationalities. Biases in financial decision making take four main forms, according to Tversky and Kahneman (1979). The second form of CPD concerns over-optimism and confidence, aversion to loss and regret, and the status quo effect. The investor who is victims of optimism bias tends to believe that what is good for him or her will inevitably happen (Bernazi, Kahneman, & Thaler, 1999). In some cases, they are conditioned by "self-fulfilling speeches" (or "performative narratives"), which give them the illusion of controlling events, according to the ancient heuristic of "magical thinking" (Skinner, 1948). This frequently observed bias explains why certain socio- economic actors persevere in situations of failure. Overconfidence, which is the most widespread form of judgment deviation, leads to the illusion that the decision-maker "understands the situation" and is capable of interpreting the trend. The third form covers emotional biases, sentimental affects and moods. These experiences have been classified by Weiner and Graham (1989) into eight classes (pride, satisfaction, joy, shame, guilt, humiliation, fear, envy), inducing different behaviours in the face of identical situations. One form finds its source in the "heuristic of affectivity" updated by Slovic et al. (1978) and based on a biased perception of risk. The contributions of emotional intelligence (EI)EI aims to understand and pilot organizational behaviors, considered irrational or inefficient, by borrowing concepts from the psychology of work and the sociology of organizations, but above all, from the psychoanalysis of management. EI is defined by Mayer and Salovey (1997) as "the ability to perceive and express emotions, to integrate them into one's own thinking, to understand and reason with emotions, and to regulate one's own emotions and those of others". Goleman's work (1995) shows that EI is an effective lever for socialization and efficient management. However, the concept of EI is controversial: intelligence is cognitive, while emotion is affective, defined as the "generic nature of pleasure, pain and emotion" (Assoun, 1997). This polysemy of EI, located at the interface of cognition and emotion, has favoured the construction of various models designed to better understand and test mental activity in its multiple dimensions: Thurstone and Spearman's multifactorial model of intelligence, Thorndike, Weschler and Lautrey's irrational factors of mental activity, Gardner's manifestations of intra- and extra-personal intelligence, Huteau's figures of intelligence... This proliferation has even led to the questioning of the notion of "bias" (Kinjal Dave, 2019). Dealing with bias and the return of trustThree types of practices can reduce - if not eliminate - the various biases or effects identified above. The first consists of framing decision making, its execution and control by means of an adapted "decision-making structure" (Thaler and Benartzi, 2004). The second is personal development. Fishoff (1982) considers that cognitive biases are generally easier to reduce than affective and emotional biases, which should logically be eliminated by experience. The third category of procedures resides in the collective ways in which people learn to make decisions. These involve hands-on training in EI, with identification of biases and experimentation with their countermeasures, and routinization of decision-making processes, particularly within communities of practice (Lave and Wenger, 1991). The efficiency of these practices conditions the return of confidence of stock market investors, company stakeholders, users of AI software, etc. Confidence generates "levels of cooperation that are much higher than those provided by the strict application of the principle of rationality" (Orléan, 1994) and is a powerful factor of personal development (by reinforcing self-confidence). However, trust is an issue for the psychoanalyst in terms of transferential relationships (Askofaré, 2009). Transfer refers to the process by which unconscious desires are actualized on objects, because of their hold, as soon as their use solicits their unconscious repetition. It most often appears as a mechanism of resistance. In the long term, the process can engender distrust, following a phenomenon of an "enantiodromic" nature according to Heraclitus. This term means running in the opposite direction, resurrected by the work of the psychoanalyst Jung (1916), to designate an immanent tendency of the unconscious. Three components of a psychic and intrapsychic nature are generally attributed to trust: conative, cognitive and affective. The first corresponds to an attitude of risk-taking in a relationship of exchange (Morgan and Hunt, 1994). The second refers to the belief in the competence of a partner or system deemed credible and reliable (Gurviez and Korchia, 2002). The third is based on recognition of the partner's benevolence or availability (Ganesan, 1994) and integrity, loyalty or honesty (Moorman et al. 1992). The insights and contributions of psychoanalysisThe call for papers is positioned in the field of Psychoanalysis & Management. Concerning psychoanalysis, we underline that its development and application have largely extended to the field of organizations, groups and institutions. Psychoanalysis, which was inaugurated by the work of S. Freud, continued and developed by his successors, has opened up new epistemological perspectives in its fields of application, particularly for the analysis of the functioning of groups, organizations, institutions and cultures. We will mention in particular the development and contributions made by the extension psychoanalysis movement (Kaës, 2013). Psychoanalysis also extends to the field of social psychology. Research in management research makes extensive reference to Lewin's work. With regard to the theme of the colloquium and for publication: It is worth emphasizing a few fundamental points brought by his insights, which we support, for information purposes, from three perspectives. There is no unbiased knowledge(s), and in this respect, psychoanalysis, as theory and practice of interpretation, of the unconscious impulse and of transferential links, challenges the disconnect between the cognitive and the emotional. Consecutively, there is no epistemological cleavage to be maintained in the exploration of thought, its genesis, its construction and its processes. Psychoanalysis can help to understand the issues and problems posed in the register of the elaboration of thought and its developments. The first step is to understand the limits imposed by algorithmic governance, the development of AI and the intensive use of digital applications, in its current and future constructions, in relation to the rights-of-way, impacts and futures ... concerning the anthropological and clinical evolution of human relationships, but also concerning the situations and practices themselves. One of the findings is that algorithmic governance has a hold on the definition of spaces of thought, its functioning and its contents, which it saturates with information. It corrupts otherness in human relations. At its limit, the subject's response is its erasure if he or she has no other alternatives. From the point of view of psychoanalysis, it is a question of symbolic violence which supports the violence already instituted by the current modalities of management, of the functioning of organizations, which is manifested by behaviors whatever their modalities and their register. The expected contributions of the conferenceThe digitalization of processes and the diversification of algorithms are guiding decision making and innovation processes within organizations - for example as a result of the development of teleworking since the health crisis. This perspective opens up new fields of research into the governance and management of organizations, extended to the work behaviors of their various actors (Sunstein and Thaler, 2008; Sunstein and Reid Hastie, 2015). The development of AI - especially machine learning and deep learning - is accompanied by new types of bias, whose nature, origins and effects are still unknown. The papers expected at the conference scheduled for 2021 will contribute to enriching reflection on the contributions of psychoanalysis, and more broadly of the sciences of the psyche and their peripheral disciplines (critical sociology, anthropology and ethnopsychoanalysis, language sciences, semiology, etc.), in the study of decisional and algorithmic biases, as well as more broadly, in the profiling of "man-machine" or "augmented man". As part of the expectations, the conference wishes to collect important contributions on the theme of the economy and management of innovation. Digital development is now proposed as the spearhead of innovation and decision making, in these different points of view, technological, organizational, managerial. This theme is a topical one. However, beyond its timeliness, it raises controversies, particularly with regard to its influence on human development. In this respect, the challenge is to show that innovation can help remove the epistemological obstacle that successive crises have highlighted. The conference proposes publication in the journal "Innovations - Revue d'Économie et de Management de l'Innovations / Journal of Innovation Economics & Management" (Revue Classée AERES, CNRS, FNEGE, ÉconLit, Scopus, ERIH Plus). The I-REMI and JIE&M Journals publish original articles and not translated articles respectively. The subjects are to be positioned on the following axis: From decisional biases to cognitive biases. The contributions of psychoanalysis in the decision-making and innovation processes of digital organizations. Publication in these two journals requires that the articles fit in with their editorial line and that the innovation is not considered as an element of context but is at the heart of the themes and methods of analysis selected. Bibliographical referencesASKOFARE, S. (2009), « Quelle doctrine du contrôle ? », Mensuel de l'École de psychanalyse des Forums du Champ lacanien, n° 44. ASSOUN P-L (1997), Psychanalyse, PUF. BENARTZI S., THALER R. (2001) “Naïve Diversification Strategies in Retirement Saving Plans”, American Economic Review, 91, 79-98. BENARTZI S., THALER R. (2001) “Excessive Extrapolation and the Allocation of Accounts to Company Stock”, Journal of Finance, 56, 1747-1764. BENARTZI S., KAHNEMAN D., THALER R. (1999), “Optimism and Overconfidence in Asset allocation Decisions”, News.morningstar.com. CALISKAN A., JOANNA J. BRYSON & ARVIND NARAYANAN (2017), « Semantics derived automatically from language corpora contain human-like biases », Science. FESTINGER L.A. (1957), A Theory of Cognitive Dissonance, Stanford University Press. GOETZMAN W., PELES N. (1997), « Cognitive Dissonance and mutual fund Investors », Journal of Financial Research, 20, 145-158. GOLEMAN D. (1995), Emotional Intelligence, Bantam. GRAHAM J.R, HARVEY C.R. (2001), “The theory and Practice of Corporate Finance”, Journal of Financial Economics, vol 60, 187-243. GREENFINCH P, « Main Behavioural Finance Concepts », https://pgreenfinch.pagesperso-orange.fr/, 2005. HEATH C., TVERSKY A. (1991) “Preferences and beleefs: ambiguity and competence in choice under uncertainty”, Journal of Risk and Uncertainty, 4, 5-28. HIGGS MJ., DULEWICZ V, (2002), Making Sense of Emotional Intelligence, 2nd ed, Windsor. D'IGNAZIO , LAUREN F. KLEIN C. (2020), « 2. Collect, Analyze, Imagine, Teach », dans Data Feminism, MIT Press. KAËS R., (2012), Le Malêtre, Paris, Dunod KAËS R., (2013), « L’extension de la psychanalyse. Introduction à quelques problèmes épistémologiques », Cahiers de Psychologie clinique, n° 40, pp. 47-69 KINJAL D. (2019), « Systemic Algorithmic Harms », Data & Society, 31 mai 2019. LAVE L., WENGER K. (1991), Legitimate Peripheral Participation, Cambridge University Press. LAMBRECHT A & TUCKER C. (2017), « Algorithmic discrimination ? : apparent algorithmic bias in the serving of stem ads », Unpublished manuscript, Massachusetts Institute of Technology MAYER ET AL. (1998), Emotional Intelligence in everyday Life, a Scientific Inquiry, Psychology Press, p.133-149. MOORMAN C., ZALTMAN G. (1992), “The dynamics of trust within and between organizations”, Journal of Marketing research, 29, pp.314-318. NISSENBAUM H., « How computer systems embody values », Computer, vol. 34, no 3, mars 2001, p. 120–119 ORLEAN A. (dir.), (1994), Analyse économique des conventions, Paris, PUF. SLOVIC P.(1978), “Analyzing the Expert judge”, Journal of applied Psychology, 53,255-263. SUNSTEIN & HASTIE R. (2015), Wiser : Getting Beyond Groupthink to Make Groups Smarter, HBR Press. THALER R.H., BENARTZI S (2004),”Save more to-morrow: using Behaviral Economics to increase Employee Saving”, Journal of Political Economy, 112 (1), 164-167. THALER R. H., TVERSKY A., KAHNEMAN D, SCHWARTZ A. (1997), “The effect of Myopia and loss Aversion on risk Taking ; an experimental test“, Quaterly Journal of Economics, 112(2), 647-661. THALER R.H, SUNSTEIN C.R.(2008), Nudge: Improving decisions about health, wealth and happiness, New haven, Yale. WEINER B., GRAHAM S. (1989), “Understanding the Motivational Role of Affect : Life-span Research from an Attributional Perspective“, Cognition and Emotion, 3 (4), 401-419. PublicationsThe best Papers will be proposed for publication in the following journals. The modalities of submission will be communicated at the end of the pre-selection to take into account the specific provisions of each journal. Please read the general conditions for submission, evaluation and participation in the conference (Call for Papers Annex, p.8).
Scientific committee
|
| Online user: 4 | Privacy |
|
