Part 2 – Late Antiquity Christian Insights for Artificial Intelligence: An anthropological approach to discerning AI’s impact on society

In Part 1 of this essay, I explained how the writing of the early Church Fathers (“patristics”) from the 4th Century CE is relevant to some of today’s important questioning about AI and society, and more specifically,  AI and faith.  This paper focuses on the question of human nature and explores the impact of AI[1] on human society through an anthropological perspective as developed by patristic scholars.

This Part 2 continues the anthropological perspective as developed by patristic scholars, drilling down on the application of the work of John Chrysotom to issues concerning the impact of AI on human society

Before tackling issues such as the composition of human nature; the mind, soul and body; the role of the nous ( gr.) in discernment, consciousness, singularity, identity, human flourishing, the meaning of work, etc. —  all  central aspects to AI and mainly to AI Ethics inquiry — I would like to cast light on two aspects that are guiding this approach:

  • In this essay, I use an anthropological methodology I identified in John Chrysostom’s works when he is analyzing human motivations, attitudes, and behaviors. It is this knowledge of the human mind and soul that I use as a research tool to scrutinize how AI concretely impacts human beings at both personal and collective levels. In the last twenty years, the patristic scientific literature has highlighted the relevance of patristic scholarship for various disciplines from psychology and cognitive sciences to economics, and now to artificial intelligence.
  • What makes these writings a treasure for the field of AI Ethics[2]? And why is this anthropological methodology and tool of analysis appropriate and pertinent to study AI, allowing us to enter into a dialogue with other methodologies used in the AI Ethics field?

While the texts under study are religious texts from Late Antiquity, they have two significant traits in common with literature on AI Ethics:  a philosophical background and a human-centered approach. These shared traits overcome any potential anachronism and converge towards universal questions about human nature.

The treasure of patristics literature for AI ethics consists in its excellent knowledge of the human soul, of the human in general. This is because these writings were authored by accomplished scholars who also reached the highest level of spiritual life.  They were masters in mystical prayer. Living in caves, they practiced metanoia (Gr. “turning back from”) and nepsis (Gr. “awake, vigilant”), meaning that every moment was a practice of virtues and a fight of passions/pathologies, made possible through attention to thoughts and vigilance of mind. Hence, they speak about human nature – body, mind and soul — out of their own experience, and their knowledge speaks to all humans from all times in all cultures.

The Chrysostomian methodology for analyzing human nature

John Chrysostom’s analysis of man’s motivations and behaviours follows a three-level perspective, best considered as three concentric circles. The human being is seen as though through a telescope with three lenses:

  1. The first level is the one looking at the human being from the perspective of her own human nature — a face-to-face between the actual human being, her thoughts and behaviours, and her human nature.
  2. The second level focuses on the human being in her relationships with others in the family environment, the work environment, and society in general.
  3. The third level sees the human being in her relationship with God.

These three levels are interconnected and influence each other; however, they address specific questions and consider different facets of the human being.

It is interesting to observe that the focus on God comes at the last level, not the first, as we may otherwise expect. Why does Chrysostom put first the analysis of human nature and its state of healthiness, even before actual behavior? Chrysostom argues that a mind and a soul that are in disorder cannot look for God. Therefore, before starting any spiritual journeyman has to get rid of bad habits, bad thoughts, to free himself from any servitude to visible things such as wealth, or to invisible things such as vainglory.

In order to analyze the impact on the human being and the nature of any activity undertaken on a regular base (for example, the use of AI tools), it is necessary to consider the impact at each of Chrysostom’s three levels individually but also interconnected.  In the 3rd installment of this paper, I will discuss more fully why this 3-level approach that Chrysostom proposes is highly relevant for AI Ethics.

But now, let’s see how Chrysostom analyzes human nature.[3]

Human Nature: Interdependence and Hierarchy of body and soul[4]

Chrysostom’s homilies reveal a man made up of body and soul with a double foundation, visible and invisible. Chrysostom’s anthropology follows a dichotomic structure, but it is different from the philosophical vision of man. Greek philosophy reveals a man made up of body and soul, and in this dualism, only the soul is immortal, while the body perishes forever. In this philosophical perspective, the body has a mostly negative role in life, and human beings must get rid of it to reach true happiness, as the body prevents humans from being happy.

Chrysostom’s anthropology is completely different.  It is a biblical one and echoes the medical conceptions of health and sickness in Late Antiquity[5]. The body is so strongly bound to the soul that each can interfere with the other’s functioning. A healthy body strengthens the soul, and bodily disorders weaken and sicken the soul. Chrysostom rejects a Manichean vision of a contemptible body, but rather underscores its importance in man’s life. In return, the condition of the soul shows through the body. The diseases of the soul spoil the face and the body, while virtues beautify it. In accordance with this vision of man, the destiny itself of the human being is considered in a biblical way, meaning that man is destined to gain eternal life, body and soul together. Their separation brought by death is only temporary and unnatural to mankind.

 

This dual anthropology is central when the AI debate about consciousness and sentience of machines turns to what I call the “digitalization of man”.  For example, when the body is replaced by avatars in a virtual world and physical functioning is overridden by digital functional equivalence, human beings cease to exist as a duality of body and mind, with significative consequences for our conception of human nature. (Again, more about this in the 3rd installment.)

An inner anthropological hierarchy

Not only does Chrysostrom posit a strong connection between body and soul, he sees a hierarchy which guarantees stability and equilibrium to this connection. The soul is superior to the body. If this hierarchy is not observed, serious disorders prevent man from being healthy and happy. The main representative paragraph is in Homily 34, 5[6]:

“For it is not the body wherein the beauty lies, but the expression,[1]and the bloom which is shed over its substance by the soul. Now then, I bid thee love that which makes the body also to appear such as it is. And why speak I of death? Nay even in life itself, I would have thee mark how all is hers that is beautiful. (…) For nothing is fairer, nothing sweeter than a beauteous soul. For while as to bodies, the longing is with pain, in the case of souls the pleasure is pure and calm[7]. Why then let go the king, and be wild about the herald? Why leave the philosopher, and gape after his interpreter?”( author’s italics)

In the paragraph above, we can see that while the body is indispensable, only when the hierarchy of soul and body is respected can man reach fulfillment. In other words, the invisible part must control the visible part of the human being for man to find happiness, to accomplish his natural beauty and excellence.

As the world is digitally transformed and AI becomes ubiquitous in our lives, this distinction Chrysostom uses between the visible and the invisible part of the human being is key to disentangling the various impacts of AI-powered technology on the human being, on her body and also on her mind. (I will also discuss this further in part 3.)

In this endless perfecting process of the human being which we find both in philosophy and in patristic studies, the nous[8] (the intellect, the natural one)  is the leading power of the mind and body that allows human beings to discern what is good and what is not, and to make righteous decisions. The nous is therefore the “headquarters” of all decision-making processes, our  “maximizing/optimizing function” for achieving human excellence.

So what does the centrality of the nous mean for AI? AI is transforming human life at a personal level, family level, professional level and collective/community level, but first and foremost, AI is changing how we think, our mind’s cognitive and affective abilities, exactly what Chrysostom was highlighting as the most important part of the human being.  The essential question is:  do we have the capabilities to measure these impacts, their length, their intensity, their potential to change human nature completely and definitively?

The next aspect of patristic anthropology even more fully reveals the importance of carefully evaluating use of AI and its impact on humans.

Human beings are noble and free creatures

According to Chrysostom, man is naturally intelligent, tender and charitable, fundamentally good. Virtue for Chrysostom corresponds to the nature of man (Homily 20. 4)[9]:

“For God, He saith, gave us understanding, that we might chase away all ignorance, and have the right judgment of things, and that using this as a kind of weapon and light against all that is grievous or hurtful, we might remain in safety. But we betray the gift for the sake of things superfluous and useless. (…) even so, if thou corrupt the mind (which hath power to put down our passions), although thou set it by a treasure, so far from doing it any good, thou hast inflicted the very greatest loss, and hast harmed thy whole soul. ”

For Chrysostom, the nous plays the role of a general or a physician.  When the eye of the body – the mind – is neglected “for the sake of things superfluous and useless”, man can only suffer and that unity in the soul/body hierarchy of human nature is damaged.  Ignorance cannot be an excuse because man is a reasonable being and he has an inner capacity for understanding[10] as well as right judgment of things — thank to the nous which is that faculty of the soul capable of making rightful judgments about things and events. It is a vital part of human nature to make decisions, enabled by the nous. Because man is a noble being capable of right judgment, he has all the capacities that allow him to be happy. But then, how to explain the evil that man does?  Chrysostom contests the idea that human beings are bad by nature or the idea that God made human beings bad. He considers evil as unnatural for mankind, in contrast to virtue, taking examples from daily life as shown in the 59th homily[11].

Because Chrysostom contends that evil is not natural to man,  the role of the will and the process of choice become central aspects of his anthropology. Passions and pathologies (of the soul first, and second of the body, from the  Gr. pathos ) are what negatively affect human discernment and judgment, first and foremost by enslaving humans. Therefore, passions have an anthropological impact – nobleness and liberty are influenced by passions and human nature itself suffers. Yet is passion a fatal flaw that man cannot change?  Chrysostom answers “no”.  In the 3rd installment I will discuss how he thinks human beings can free themselves from this influence.

This issue of whether humans in fact have freedom of choice is another critical question in our current debate about AI’s impact on human flourishing.  Chrysostom argues that as humans are by nature intelligent and noble creatures, they can fight against and correct wrong decisions made owing to human inclinations where the nous couldn’t decide what is right. But increasingly, the power to manipulate human decision-making through collection and analysis of vast amounts of personal data calls into question the very idea of free will.

This issue of free will is at the root of how artificial intelligence interacts with the nous and the ultimate question of whether human beings can still flourish in a world where AI is ubiquitous.  If the answer to that question is yes, then how?  If no, then why not?

Part 3 to come will consider some possible paths to answer these questions.

[1]  While it is difficult to define AI, here is a good resource among many others: AI-Definitions-HAI.pdf (stanford.edu), Defining AI | One Hundred Year Study on Artificial Intelligence (AI100) (stanford.edu). For events on AI, among the best ones are: NeurIPS: Conference on Neural Information Processing Systems; AAAI Conference on Artificial Intelligence; AI World Congress 2020; The AI Summit; the IEEE Conference; to name only a few.

[2] The concept and the area of research of AI Ethics may suggest that there is also AI without ethics. All AI is AI Ethics because there is no neutral AI. As any of our actions uphold a preference for certain outcomes based on our values, the same is true for AI and the algorithms and models behind it. All AI-based decisions have a rule of maximising, optimisation which intrinsically values a variable higher than the others in order to reach the desired output.

[3] I use here the corpus from Chrysostom’s Homilies on Matthew, Homilies on Genesis, and Baptismal Catechesis.

[4] This text is drawn from my previous work on John Chrysostom’s Homilies.

[5] Christoph Markschies, ‘Demons and Disease’, SP 81 ( 2017), 11-36.; Ellen Muehlberger, ‘Theological Anthropology and Medecine: Questions and Direction for Research’, SP 81 ( 2017), 37-50.; Stefan Hodges-Kluck, ‘Religious Education and the Health of the Soul according to Basil of Cesarea and the Emperor Julian’, SP 81 ( 2017), 91-108.; Jessica Wright, ‘John Chrysostom and the Rhetoric of Cerebral Vulnerability’, SP 81 ( 2017), 109-126; Chris de Wet, ‘ Gluttony and the Preacher’s Diet: Regimen, Obesity, and Psycho-Somatic Health in the Homilies of John Chrysostom’, in Wendy Mayer and Chris de Wet ( eds), (Re) Visioning John Chrysostom ( Leiden and Boston); Wendy Mayer, ‘The Persistence in Late Antiquity of Medico-Philosophical Psychic Therapy’,  JLA, 8.2 ( 2015), 337-51.

[6] Œuvres complètes de Saint Jean Chrysostome, Français-Grecque, Edité par l’Abbé Bareille, Paris, 1866-69, Vol. 12, p. 165.

[7]  j.s. mill, L’utilitarisme. Essai sur Bentham, Paris, PUF, 1998, p. 33, 37, 90, does  the same analysis as Chrysostom does in this paragraph.

[8] nous | Greek philosophy | Britannica

[9] Œuvres complètes de Saint Jean Chrysostome, Français-Grecque, Edité par l’Abbé Bareille,  Paris, 1866-69,Vol. 1, p. 595, see also  page 593.

[10] Interesting to remember here that John Locke was writing his Treatise on human understanding as an answer to a question raised by his friends after a meeting where the subject was “What are the bases of the principles of morals ?”».

[11] Œuvres complètes de Saint Jean Chrysostome, Français-Grecque,  Edité par l’Abbé Bareille,  Paris, 1866-69,Vol. 12, p. 469-477.

X
%d bloggers like this: