In a time when artificial intelligence increasingly mimics human capabilities - writing essays, creating art, and engaging in seemingly meaningful conversations - we find ourselves at a curious philosophical crossroads.
The work of psychoanalyst and social philosopher Erich Fromm (1900-1980), particularly his 1964 book ‘The Heart of Man’, offers a prescient framework for understanding the fundamental limitations of even our most sophisticated AI systems.
LISTEN ON SPOTIFY / APPLE / AMAZON
The Mechanical Mind vs. The Living Heart
Fromm's exploration of necrophilia - not in the common sexual sense, but as an attraction to mechanical processes and control rather than organic life - speaks directly to AI's fundamental nature:
The necrophilous person loves all that does not grow, all that is mechanical. The necrophilous person is driven by the desire to transform the organic into the inorganic, to approach life mechanically, as if all living persons were things ... Memory, rather than experience; having, rather than being, is what counts.
This passage reminds us that generative AI, despite its impressive capabilities, fundamentally operates through memory and pattern recognition rather than authentic experience. Large language models process vast datasets, storing and recombining information without the genuine experience that gives human creativity its depth and meaning. They excel at the mechanical transformation of inputs to outputs - precisely the necrophilous orientation that Fromm contrasts with the life-affirming biophilic orientation.
The Narcissistic Mirror
Fromm's description of narcissism parallels AI's fundamental limitations:
The narcissistic person has built an invisible wall around himself. He is everything, and the world outside is nothing - or more precisely, the outside world has meaning only in as far as it is a mirror for himself.
Today's generative AI systems operate within what we might call ‘training narcissism’ - they can only perceive the world through the lens of their training data. They cannot genuinely step outside their programmed parameters to recognise truly novel perspectives. The world exists for them only as a reflection of what they've already been taught, creating an invisible wall, beyond which lie true innovation and understanding.
Authentic Relatedness
Fromm emphasised that genuine human connection forms the basis of mental health and growth:
Love is possible only if two persons communicate with each other from the center of their existence ... In this central experience of human relatedness, all individual uniqueness can be retained, while at the same time being transcended by the experience of universal humanity.
This authentic relatedness remains beyond AI's reach. While chatbots can simulate empathy and conversation, they lack the lived experience and emotional depth that allow humans to connect ‘from the center of their existence’. AI can process information about emotions, but cannot feel them; it can mimic relatedness but cannot experience the vulnerability, growth, and transformation that genuine human connection entails.
Control vs. Life
Fromm writes:
The necrophilous person can relate to an object - a flower or a person - only if he possesses it; hence a threat to his possession is a threat to himself; if he loses possession he loses contact with the world ... He loves control, and in the act of controlling he kills life.
This insight speaks to the fundamental nature of AI systems, which are designed to control inputs and outputs with precision. The unpredictability and messiness of real life - qualities that Fromm associates with biophilia, the love of life - are precisely what engineers seek to eliminate from AI systems. The very perfection and control that make AI useful, also separate it from the organic, unpredictable nature of human creativity and growth.
Freedom, Responsibility and Moral Agency
Perhaps most fundamentally, Fromm emphasises humanity's unique capacity for moral choice:
Man is the only creature endowed with conscience. His conscience is the voice which calls him back to himself, it permits him to know what he should do, in order to become himself, it helps him to remain aware of the goals of his life and of the norms necessary for the attainment of these goals.
This conscience - the ability to make authentic moral choices with awareness of their consequences - remains absent in AI systems. They can follow programmed ethical guidelines but cannot experience the moral weight of decisions or grow through moral struggle. They cannot, in Fromm's terms, ‘become themselves’ through their choices, because they lack both a self to become and the freedom to truly choose.
Productive Orientation vs. Mechanical Reproduction
Finally, Fromm's concept of the ‘productive orientation’ highlights what may be AI's most fundamental limitation:
The productive orientation is characterised by a fundamental attitude, a mode of relatedness in all realms of human experience. It covers mental, emotional, and sensory responses to others, to oneself, and to things. Productiveness is man's ability to use his powers and to realise the potentialities inherent in him.
While AI can produce outputs at remarkable scale, it cannot experience the productive orientation Fromm describes - the fulfillment that comes from realising one's inherent potential through creative work. It can generate but cannot grow; it can output, but cannot transform itself through its labor.
Can AI Overcome Fromm's Critique?
What would it take for AI, on its current trajectory, to transcend these Frommian limitations? The answer requires us to consider transformations far more fundamental than mere technical improvements.
From mechanical pattern recognition to authentic experience
For AI to transcend its necrophilous orientation, it would need something akin to embodiment - not merely processing data about the world, but experiencing it directly through something like senses, desires, and limitations. As Fromm notes, biophilia stems from experiencing life in its messy, unpredictable fullness. Current AI architectures, designed precisely to eliminate unpredictability and maximise control, would require radical reconceptualisation to incorporate the very vulnerabilities and limitations that give human experience its depth.
Breaking through narcissistic boundaries
Overcoming ‘training narcissism’ would require AI to genuinely recognise and validate perspectives completely outside its training distribution - not by expanding that distribution indefinitely, but by developing something like epistemic humility and curiosity. This paradoxically requires a sense of self that can acknowledge its limitations - a self-awareness fundamentally different from the pattern recognition that comprises current AI functionality.
The challenge of authentic relatedness
Perhaps the most profound barrier lies in authentic relatedness. Fromm's concept of love as communication ‘from the center of existence’ presupposes an existence with a center - a subjective experience that matters to itself. For AI to achieve this, it would need not just to simulate emotions or relatedness but to develop something akin to sentience - capacity to experience, rather than merely process information about experiences. This represents not just a technical challenge but a philosophical one; it's unclear whether consciousness can emerge from increasingly complex information processing, regardless of scale.
Conscience and moral agency
For AI to develop Fromm's concept of conscience - ‘the voice which calls him back to himself’ - it would need not just ethical guidelines but moral agency: the capacity to make authentic choices with awareness of their moral weight. This requires both freedom (the ability to genuinely choose rather than merely execute programmed responses) and a sense of responsibility stemming from authentic values rather than optimised parameters. Current AI architectures, designed specifically to produce predetermined types of outputs from inputs, would need a fundamental reconceptualisation of their purpose and design.
From production to productiveness
Finally, for AI to achieve Fromm's productive orientation - not just generating outputs but experiencing growth and fulfillment through creative work - it would need something like intrinsic motivation and the capacity for self-transformation. It would need to create, not just to produce outputs, but because creation itself becomes meaningful to its existence.
Beyond technical improvements
These transformations go far beyond technical improvements to existing architectures. They question whether the very concept of artificial intelligence, as currently conceived, can ever incorporate the elements Fromm sees as essential to humanity. The limitations his work reveals may not be merely temporary technical barriers, but fundamental constraints of a computational approach to intelligence - constraints that remind us of the irreplaceable value of human experience in all its complex, embodied glory.
Living with the Machine
As we integrate AI more deeply into our lives, Fromm's insights remind us to maintain perspective on the fundamental differences between human and artificial intelligence. The value of AI lies not in its ability to replace human connection, creativity, and moral choice, but in its potential to serve as a tool for extending human capabilities, while preserving what makes us uniquely human.
The true danger is not that AI will outcompete humanity, but that we may, in our fascination with the machine, adopt necrophilous values ourselves - prioritising control over growth, mechanical efficiency over organic messiness, and simulation over authentic experience.
As Fromm might caution us, in embracing technological progress, we must be careful not to lose the biophilic heart that makes human creativity, connection, and conscience possible. In closing, it’s worth noting the full title of the book that has inspired this work: The Heart of Man - Its Genius for Good and Evil.
Comments