Artificial Intelligence (AI) - Ethics and Meaning from a Jungian Perspective
- Nicholas Toko
- Dec 14, 2023
- 5 min read
Updated: Dec 18, 2023
The nature of psyche and implications for AI
Topography of the Psyche.
From Jung's Indispensable Compass. Navigating the Dynamics of Psychological Types by James G Johnston.
‘Artificial Intelligence: Ethics and Meaning from a Jungian Perspective’ is an editorial collaboration between Nick and Margot dedicated to exploring Artificial Intelligence from a uniquely Jungian psychological perspective and with a specific focus on the meaning and ethical implications of artificial intelligence technology on individuals, groups and society.
Nicholas Toko is a Jungian Analyst-in-training at ISAP Zurich. He has an MA in Jungian and post Jungian Studies from University of Essex, MA in Human Resource Management and an International Business Studies degree from Brunel University. He recently completed an AI Programme at Säid Business School, University of Oxford.
Margot Estabrook Stienstra, Lic. phil. is a Training Analyst at ISAP Zurich /AGAP, IAAP and licensed Psychotherapist (ASP) in private practice in Zurich. She holds a M.I.M. in International Management from Thunderbird School of Global Management and shares a keen interest in Artificial Intelligence.
Margot's profile
Before shedding light on ethical implications per se, it is important to lay some groundwork about our chosen context of C.G. Jung’s Analytical Psychology as a framework from which we can elicit a sense of Jung’s - and our own- ethical standpoint. Furthermore, this context offers parameters for our exploration of the nature of AI and the unfolding reciprocal relationship with humankind.
So we begin by establishing the complementary relationship of what we generally think of as consciousness to what Jung called the #collectiveunconscious. This relies on the capacities of our ego and its complement in the #unconscious, the archetypal image of #Self.
The ego’s domain is personal, subjective, and one of perceived urgency, with its necessary tasks of helping us to orient daily to the world, gleaning and prioritizing relevance, usefulness and efficiency, as well as supporting our internal and external adaptation through regulation of the antennae of our emotions. The ego is often experienced as self-sufficient and self-reliant, but without communication and ongoing relationship with its more ancient evolutionary and ubiquitous root, the Self, our well-meant adaptational choices and decisions are ultimately one-sided and lead us to displaying neurotic behaviors that lack self-insight or the ability to truly relate to others. Others are quick to see our (autonomous feeling-based) #complexes, whereas for ourselves they are en-shadowed. (I’ll unfold more on this next month.)
Jung emphasized that the aforementioned one-sidedness to which we are prone even when determined to live consciously, requires #compensation; our conscious perceptions can be more fully realized and with less animosity and loss of energy when tempered with more longstanding wisdom of the archetypal Self, which Jung identified as an ordering principle governing essential images for the making of meaning.
The crucial self-regulating power of the #psyche or soul sets the tune and rhythm for the
day-to-day “dance” between a determined ego, with its tendency to claim territory and stake a flag pronouncing convictions, and the more tempered, fluidity of the Self.
W.B. Yeats expressed this well in his poem “The Second Coming”:
The best lack conviction;
The worst are full of passionate intensity…'
This micro (daily) and macro (over a lifetime) dynamic Jung named the #individuation process.
So the act of conceiving of and inventing Artificial Intelligence was achieved primarily through the human ego and so risks limited vision that Jung called Shadow, to be further discussed next month. So as we continue to discuss the potential and limitations of AI, we must keep in mind that it may well be inherently flawed or at least incomplete by virtue of being a human invention constituting both conscious and unconscious intentions.
More concerning is that AI is designed to have its own agency and autonomy. It has in this sense its own ego. But what would the unconscious have to say about this supposedly conscious gestalt? What is currently beyond our current seeing, knowing, understanding that could be - and most likely is - consequentially affecting “our” invention? Is there any collective evolutionary wisdom weighing in through the portal of the unconscious? And how similar is AI really to ourselves if we consider the functioning of our body, and in the brain, its instinct-based “reptilian” stem, the limbic region that includes our emotional, learning, motivational capacities, and our neo-frontal cortex with its more rationally based executive function capabilities, such as reasoning, risk evaluation, and forecasting?
One quickly realized that the unconscious is inherently cryptic. At best we can manage only glimpses - and for many who can easily justify a more rational orientation, what is out of sight and mind appears unfathomable, inefficient and hard to take seriously. Even many neuroscientists find less conscious human behaviors - human irrational grappling with emotion for instance, to be a nothing short of a bloody nuisance!
Let’s now take a few minutes to consider Jung’s psychic map (See Murray Stein’s fine book, Jung’s Map of the Soul, listed below) and make some initial assessments of whether parallels exist in AI. Readers may find it helpful to make a Venn diagram as we consider comparisons and contrasts…
Guiding Points and Questions for readers’ consideration:
How might each of the concepts shown in bold above be reflected in the context of AI, i.e. where can we “locate” each in the context of AI?
The unconscious of the AI engineer could influence the development of the algorithm in positive or negative ways. The psyche or totality of the AI engineer's personality could also have an impact on the algorithm, for examples, does the AI engineer project the 'Self' to the algorithm, how would this manifest in the AI machine?
Can we observe any obvious parallels?
There is no single, all-encompassing definition for AI, but rather a wide range of definitions which attempt to define what AI truly means. According to Russell and Norvig (Artificial Intelligence: A Modern Approach 2016:1), the various definitions of AI can be categorised into four dimensions AI: thinking humanly, acting humanly, thinking rationally, acting rationally.
The pursuit for transformation is a common feature in psychoanalytic and AI work. So you might say there is a pursuit for psychological transformation in psychoanalysis and a pursuit for technological transformation in AI.
Jungian psychoanalysis aims to transform the perspective, attitudes and behaviour of an individual, or to achieve psychological wholeness, totality of the psyche or the Self.
Digital technology aims to transform AI by developing machines that can think humanly, act humanly, think rationally, and act rationally, technological wholeness or the singularity.
It is also worth considerating the individuation process and it's parallels with the AI engineer's pursuit for technological transformation, truly intelligent AI on a par with the human being.
Can we note any complementary or compensating relationships within AI?
Think about the relationship between the analyst and analysand. Is there a reciprocal relationship between the AI engineer and the algorigthm within the neural network of an AI machine.
Jungian psychoanalysis takes place within a dialectical relationship between analyst and analysand (a client or patient in a psychoanalysis), and has for its goal the analysand's movement toward the 'Self" or psychological wholeness. By dialectical relationship, I mean the analysand and analyst's conscious and unconscious minds are in relationship within and between each other.
The interaction is as follows:
The conscious relationship between the ego of the analysand and the analyst also known as the conscious relationship.
The relationship between the unconscious mind of the analysand and the conscious personality of the analyst.
The relationship between the unconscious mind of the analyst and the conscious personality of the analysand.
The relationship between the unconscious mind of the analysand and the unconscious mind of the analyst.
The interaction between the AI engineer and the algorithm follows a similar structure:
The conscious relationship between the ego of the AI engineer and the AI algorithm can also be perceived as the conscious relationship. The potential origins of intended bias.
The relationship between the unconscious mind of the AI engineer and the algorithm. The potential origins of unintended bias.
The relationship between the conscious mind of the AI engineer and the neural network, the 'unknown' realm of machine learning.
The relationship between the unconscious mind of the AI engineer and the neural network.
Both interactions are illustrated in the diagram below. The dynamic interactions between the actors represented by the vectors.
A complex tangle of dialectical relationships in both scenarios, and each has for its goal the achievement of transformation, specifically, Jungian Psychoanalysis has as its goal movement towards the Self, or psychological wholeness, and AI has the goal of achieving the Singularity, or technological wholeness.
The transformation of the personality requires coming to terms with the unconscious, its specific structures and their dynamic relations to consciousness as these become available during the course of analysis (Jungian Analysis, (ed) Murray Stein, 1995). The transformation of the algorithm requires coming to terms with the neural network in which many small artificial neurons are connected in complex networks.
The process of transformation is not a linear process, there are ups and downs, successes and failures, in psychoanalytic and AI work. Machine learning and neural networks has led to the big leaps in technological innovations that we see today e.g. ChatGPT but like psychoanalysis, change is incremental and small modifications to the personality and algorithm are made along the way. It is a long term and far-reaching exploration of the personality and AI.
Which appear to be most relevant comparisons / contrast when evaluating pros and cons of AI?
All of the concepts have some utility to explore the pros and cons of AI. The AI engineer's personality, their conscious and unconscious relationship with the algorithm and ultimately, the question of meaning of AI work can be explored through the lens of analytical psychology.
What ethical questions come to mind as one contemplates what Jung called “the reality of the psyche”?
The autonomy of the psyche and its contents raises questions about what is projected on to the AI algorithm, is it positive or negative? How then does the algorithm impact individuals, groups and society?
Your own (Readers’) questions:
What questions come to your mind?
As both Jung’s Analytical Psychology with its conceptualization of the human psyche, and Artificial Intelligence are highly complex, it may help if I express the above slightly differently. Readers please make note of any intellectual and emotional responses that may arise in your considerations:
A. We begin with the notion of the Psyche or Soul: Most succinct and relevant for now is the dual nature of the psyche as Jung understood it; it is comprised of consciousness, and both the personal and collective unconscious. As unconsciousness deepens it becomes less subjective and more objective.
B. A further important extension (not mentioned above) would be what Jung called the Psychoid. In brief, this term gives us a starting point to reference the borderland or boundaries of the psyche, with the body, the physical world beyond the body, as well as the domains of what is generally referred to as “spirit.”
C. In the Western world one references an “I” as the center of our ego-consciousness, through which we more “knowingly” perceive the world, orient ourselves, and make adaptive decisions.
D. Yet by its very nature, keeping us on task, attending to what appears to be most relevant and practically or rationally “efficient,” it is typical to forget that we are more than what we consciously chose to focus on at any given time. At the centre of our larger Self is the instinctual well-spring of all archetypal sense-making and its images from the unconscious that have evolved over millennia to inform and bring order and structure to our conscious lives and “choices,” even though we may be unaware of their profound influence.
E. When turning our ego’s self-concept towards social adaptation, presenting what Jung called its Persona, only part of ourselves is absorbed in this necessity; the rest remains less- or un-conscious, and thus contains rejected or as yet unknown attributes of ourselves, which Jung called the Shadow.
F. Central to Jung’s Analytical Psychology is that the Shadow -and deeper, cultural and collective parts of the unconscious have a reciprocal, compensatory relationship, acting on us and influencing our attitudes and behavior, often very apparent to others but not to ourselves! You can begin to see the ethical implications in the choices we make of which we are unconscious! It is helpful -Jung saw this as an ethical imperative- to develop the habit of keeping the psychic reality of both in body-mind as best we can.
G. Jung identified deeper strata of the unconscious he termed anima and animus, but these can be examined in more detail as we proceed.
Recommended reading:
Murray Stein’s Jung’s Map of the Soul
Michael Wooldridge’s AI: Everything you need to know about the coming AI
James G Johnston's Indispensable Compass. Navigating the Dynamics of Psychological Types
Copyright
© 2023. Authors: Margot Estabrook Stienstra and Nicholas Toko. All Rights Reserved.
Comments