On Tuesday October 1st, the Canadian Council of Churches (CCC) and the World Council of Churches (WCC) hosted the webinar “Theology and Artificial Intelligence: Systematic and Denominational Perspectives”. The aim was to analyze challenges posed by generative AI and transhumanism for theological aspects such as the personhood and the image of God, the meaning of the Incarnation, and the Trinity.
The webinar was initiated by Father Moussa Wahib of the Coptic Church, who is Co-Chair of the Faith and Science Group of the WCC. Next, Pastor Peter Noteboom, General Secretary of the CCC, and Dr. Manoj Kurian, director of the WCC Commission of the Churches on Health and Healing, welcomed and thanked the attendees. After that, five presentations were given by experts, followed by a discussion, under the moderation of Rev. Dr. Peter Kuhnert from the WCC.
The video of the webinar is available at: https://youtu.be/GOSpDI9Phgg
Artificial Ingelligence: Perspectives
Rev. Dr. Gayle Woloschak (Northwestern University)
- Machine learning is the process of using large data to improve how machines ‘learn’ (improve performance on some datasets) for specific tasks such as email filtering, speech/pattern recognition, and analyzing scientific datasets.
- AI leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind, which we can find in applications such as facial recognition, ChatGPT, voice assistants, and others.
- Transhumanism discloses visions of mind out of body to upload to computer chips or even to particular boxes to be part of a hive mind.
- Unique features in persons (not present in AI) are: development of culture, language for relationship purposes, fulfillment of being, and responsible agency, among others.
- Two additional features that we found only in humans are incarnation and creativity. The first is important, and church fathers such as Athanasius the Great and Gregory of Nyssa defended it. The last one can be conceived as a gift to humans as an image of God.
- AI technology is a product of unique human creativity, while AI itself is not creative in the same sense that humans are. Indeed, AI technology is also a product of unique human creativity, even though it is also prone to biases given the input data.
An Impulse on AI
Bishop Dr. Heinrich Bedford-Strohm (World Council of Churches)
- The crucial problem is the extent technological production aimed at enhancing humans is reconcilable with the conviction that human beings are created in the image of God.
- From a biologistical view (like that Iuval Harari claims), human beings are seen as a series of algorithms. However, this is quite different from the Christian view, in which God gives us dominion over the works of his hands, but we still are lower.
- As it was pointed out by theologian Karl Barth, human beings can only be understood if they are interpreted from the humanity of Jesus. Then, this framework reveals two crucial features in humans: relationality and vulnerability.
- Suffering and imperfectness need to be included in the image of human beings; then AI is a means to alleviate suffering and not to perfect us.
- The Bible passage about the tree of life that is not allowed to eat from (Genesis 3) is a symbol of God’s loving will for protecting us from striving to be eternal (aspiration that is common in several AI fantasies).
- The conclusion is that AI does not have religious significance, but it is only an instrument to assist and help other people with responsibility.
Orthodox Dimensions in AI and Humanity
Very Reverend Dr. Aristarchos Gkrekas (National and Kapodistrian University of Athens)
- From an Orthodox perspective, Genesis 2:7-9 gives an account of the beginning of the relationship between the Trinitarian God and humanity.
- God created humanity in “His image”; that is, the rational and autonomous, soul and body as a unit, and “likeness”, or the best possible likeness of virtue through the life in Christ.
- Christ is the incarnation of the Word of God. Then, as Church is the body of Christ, human beings can experience theosis-deification in the life of Church to finally receive the gift of salvation and immortality.
- For transhumanism, assisted by an apocalyptic view of AI, a digital reality would offer human beings unlimited joy and well-being, centered in a paradise and a technological Eden.
- Under the above framework, (trans-)humans will be able to download their consciousness into machines, resulting in enhanced mental abilities and immortality via unbounded replication and physical improvements such as self-healing, flexibility, fast computation, etc.
- Transhumanism deifies AI by considering a possible superintelligence that would have much greater perceptual ability and knowledge than human beings. Then, this view seems to be a new form of gnosticism, a digital or algorithmic gnosticism.
Roman Catholic Theological Reflections on AI: Dignity, Incarnation, Relationality
Dr. Cory Labrecque (Université Laval)
- Following Pope Benedict XVI, we need to consider that technology is not neutral because it reveals humankind’s aspirations towards development to overcome material limitations.
- Technology must be conceived as a tool to assist humankind. If technology becomes more important than human users such that we overvalue the instrument and praise functionality, the risk of devaluating human dignity emerges.
- For transhumanism, human features such as the inevitability of aging and death, fluctuating intellectual capacities, and emotional fragility, among others, are seen as shortcomings.
- However, from a Christian perspective, human limitedness has value. A paradigmatic example is the Resurrection narrative, in which Jesus is still wounded after his resurrection, i.e., the risen Jesus would be recognized by the marks of a limited body.
- Regarding the injection of AI technologies in human relationships (for instance, companion robots and virtual avatars of deceased loved ones), there are several questions.
- In any case, any discussion about AI in relationships should address the benefits and limitations of using these technologies. In addition, these problems also invite us to revisit traditional theological ideas about care and caregiving, what constitutes human personhood, and what it is about us that we must strive to preserve.
Artificial Ingelligence: A United Church of Canada Perspective, A Few Thoughts
Rev. Dr. Tracy Trothen (Queen’s University)
- AI technologies in relationships not only have positive physical effects but also emotional, cognitive, moral, or even spiritual effects. There are studies about chatbots, such as Replika, that have shown they have positive effects, making people less lonely.
- To determine how far we should go with AI depends on what we mean by normal; however, the line between normal and beyond is blurry. Heart surgeries initially were shocking due to the sacred meaning that had that organ, but today they are normal. In the future, this would be the case with AI agents as chatbots to create avatars of deceased loved ones.
- The difficulty is to encounter AI agents as not us and not God, because they are like us and remind us of some godlike qualities. At the end, this problem requires us to stand back from ourselves and really want to get to meet and know each other.
- A more relational approach also could be useful to go beyond apocalyptic or utopic visions of AI, usually popularized in movies such as “Terminator” and “I, Robot.”
- As it was shown with the ELIZA effect in the 1960s, we have a tendency to anthropomorphize AI models, mainly motivated to want to understand based on firm bounds.
- If AI technologies have positive benefits, they can help us to care for each other (for instance, in facilities, hospitals, or even just in our homes). However, the challenge is not to give up our social responsibility to care that the Church holds, nor a relational life.
Discussion (Questions and Answers)
In the discussion, several (not necessarily convergent) ideas, were pointed out.
1. It would be important to move beyond apocalyptic AI views, to seek forms in which AI could enhance, in relationality terms, care for each other without ignoring our duties. Moreover, a useful vision of AI is one that allows us to have adequate discussions in civil society about its possibilities, limits, risks, and also regulations. However, we have to be aware that data that feeds these algorithms can be biased, which makes it difficult to achieve regulations.
2. Another important point is how we should understand “perfect” in the context of AI technologies that seek to “perfect” us. Firstly, for Christianity, the meaning of “perfection” is a certain kind of moral perfection (Matthew 5:48), being different from the meaning used in transhumanism. On the other hand, from a more practical approach, we could mention that in medicine, several applications of AI seek to achieve better diagnoses and to make physicians better, but there are some things that need to be supervised by human expertise. In any case, from a social justice perspective, we should also ask what AI medical applications are accessible to everyone.
3. Regarding fake images and videos made using generative AI, it would be important to go in a direction in which we can deal with this problem, which has several societal and political implications. We need to draw on resources to check media files to make sure they were generated by AI or not. These tools are not perfect, but making those available easily and rately to everybody, or even regulating videos that are posted, would be a big step forward.
4. Aspects in which the image of God resides are creativity, love, and relationality. Even more, some panelists claimed that some of these features are unique to humans. Moreover, the possibility of AI replacing human clerics was addressed, entailing that, from several true stories shared by some of the panelists, it seems that the best option would be to find a balance. It would be valid (and even recommendable) to use robots for assistance but not for replacement.
5. Ideas about how to deal with possible future crises arising from technology also were addressed. Firstly, it must be obligatory to include failure-friendliness in technology so that we can anticipate if mistakes are made. Moreover, given the potential negative implications on a political and worldwide scale of these technologies, we need to ask what the responsibility of faith communities is. It is known that churches are a strong source of ethical frameworks; however, it could be important to find stability by considering different ethics such that we can define a more general ethics that is able to apply and interpret through human rights and international law.
6. Finally, panelists analyzed the impact of AI on the environment. We know that huge data farms that have an extensive carbon footprint are necessary for AI. However, we also know that important AI applications to solve environmental problems exist. Here the recommendation is to search for a balance and to evaluate advantages and disadvantages of AI in terms of manufacturing for AI use and development.
Reflection
This webinar was a very refreshing initiative to discuss some of the current AI issues. Many of us appreciate it. Here I emphasize the importance of keeping these events in online format since people from other countries, and in particular from the Global South, can attend. Moreover, given the global and cross-sectional implications of AI technologies, I also congrats on the importance of discussing AI in an ecumenical and multi-disciplinary setting.
Now, from my particular perspective as a Christian astrophysicst working in the Global South (Mexico), I must draw attention to the fact that, for the sake of promoting a pluralistic perspective, some crucial topics should not be absent in discussions about AI.
- Firstly, the importance and relevance of democratizing AI technologies. Here movements such as Open Science (including aspects such as open data, open source codes, etc.) are relevant because they help to advance under the lens of social equity. Then, one crucial question to ask is: How could we theologically interpret and promote this turn in AI technology development and fundamental research?
- The necessary ethical analyses about how transnational AI megacorporations are prone to keep power structures that, in many aspects, marginalize certain regions for the wide benefits obtained from AI. We live in an age in which there is a huge power imbalance between the Global North and the Global South, but, at the same time, we are more aware of the limits of modernity and the relevance of plurality and relationality in the world.
- For addressing problems as those mentioned above, it is really important to avoid, at all costs, faulty, sloppy, and incomplete knowledge generation under neo-colonial schemes such as “parachute science.” Any international discussion about AI must consider a wide spectrum of experts in terms of their ethnicities and, particularly, those who do research in countries from the Global South.