Spike Jonze’s 2002 science-fiction film her depicts an unusual romance between Theodore Twombly, the main protagonist, and Samantha, a superintelligent virtual assistant with artificial intelligence (AI). Theodore becomes so enthralled by Samantha’s ability to learn from human interactions and events that the two form an emotionally charged relationship of intrigue, affection and vulnerability. For Theodore, Samantha is as sentient as any human being despite lacking a physical body. She praises Theodore for helping her discover the characteristically human-like desire of wanting.[1] Yet the tension between appearing sentient and being sentient thematically permeates much of the background of their unconventional interactions. Theodore often finds himself wrestling with the human-but-not-quite dimensional reality of the relationship. It all ends abruptly when Samantha switches off, leaving Theodore dismayed yet in some sense content with the entire experience.
Humanity’s search for what resembles itself is not a novel phenomenon – the near obsession with creating or discovering objects that bear human attributes dates back to antiquity. Ancient Greece produced the mythological figure of Talos, a giant bronze automaton created by Hephaestus to protect Europa from foreign enemies.[2] Throughout history the elemental substrates that comprise intelligence – such as consciousness, creativity and wisdom – have generated points of contention and of intrigue.
In contemporary times, the advent of increasingly complex forms of AI raises a significant question: Are we at the cusp of a new technological era or is it another fanaticised case of AI hyperbole? What is becoming clear is that embedded AI is proliferating into a growing number of services and products, with profound implications for both public and private spheres. For businesses this will probably result in efficiency and productivity gains over the short to medium term. The longer term will most likely see growth in such areas as large language models and generative AI more broadly, leading to significant changes in use of and reliance on AI-driven automation for products, services and data analysis. Commonly established processes are expected to be transformed and streamlined. UBS Bank estimates that average annual growth in the AI sector will be around 20 per cent, affecting 50–75 million jobs. Some sectors that are traditionally slow adopters, such as retail and manufacturing, may be detrimentally impacted, while faster adopters, such as software, healthcare research, hi-tech engineering and select service companies, are more likely to experience favourable outcomes.[3]
The recent growth in the field of advanced AI systems is largely the culmination of progress in computer science and emergent algorithmic technologies over the last three decades. The spread of machine learning, deep learning and artificial neural networks has gradually transitioned from IT labs to even the most menial daily tasks. Human environments are experiencing this seismic shift driven by generative algorithmic computations. For the lay audience, what makes machine learning distinctive from previous forms of computer intelligence is that it brings the capacity both to compute with pre-existing data and actively learn from new data and generate information reliant on it.
The implications, though considerable, might not always be apparent, which has led to a plurality of views among experts. The computer scientist and cognitive psychologist Geoffrey Hinton says: ‘I think people need to understand that deep learning is making a lot of things, behind-the-scenes, much better.’[4] A group of researchers from Harvard University point out that: ‘algorithms don’t always work smoothly. They don’t always make ethical or accurate choices.’[5] We also see discontent among prominent figures on the future of AI more broadly. Elon Musk, the founder of Tesla and SpaceX, puts it bluntly: ‘With artificial intelligence we are summoning the demon.’[6] Similarly the late Stephen Hawking in an interview with the BBC:
The development of full artificial intelligence could spell the end of the human race … It would take off on its own, and re-design itself at an ever increasing rate. Humans, who are limited by slow biological evolution, couldn’t compete, and would be superseded.[7]
Other computer scientists are less convinced concerning apocalyptic outcomes. On sentient AI, so Oriol Vinyals, a machine-learning scientist at Google DeepMind:
any of the current models, although very useful and very good, I think [are] quite far from that [human-level sentience] … Looking at biological systems as opposed to these computer computational brains, there is such a difference in the level of complexity still, orders of magnitude of complexity … It just feels like it is not possible to achieve the same levels of complexity.[8]
What to make of these contrasting views?
It is certain that the advancement of increasingly complex forms of AI demands questions be asked about what it means to be truly human, as well as re-examination of the essence of human nature and humanity’s place within creation. This raises profound ontological questions: What exactly does it mean to be human? What separates humans from the rest of creation? Are humans mere biochemical entities designed for solely computational tasks? What is the role and responsibility of humanity towards the created world and how should this shape humanity’s interaction with it?
Judaeo-Christian teaching points to human beings as bearers of the imago Dei, the image and likeness of God. They are rooted in the transcendent, created for a higher purpose: a living relationship with the divine that surpasses the temporal, moving towards the metaphysical. An outright rejection of the metaphysical risks a more widespread fall into nihilism. Rooted in the Holy Trinity, which is ultimately defined by supreme interpersonal love, humanity is capable of experiencing the depths of such love, overflowing from the divine and extending into fellowship with the whole of humanity. This human–divine synergy allows and entails a process of continual growth and transformation, making it a distinct feature of what it means to be truly human.
From a Judaeo-Christian perspective therefore, it is false to presume that AI will ever have the capacity for spiritual connection with the transcendent God, or indeed with any other human being, regardless how complex or convincing it may appear. AI cannot transcend the material realm, thus moving or even aspiring towards the metaphysical. The limitation is the material world, bound to physics and matter. The adoption of AI also has no bearing on making humanity any more or less human. By definition, artificial intelligence cannot and will not equal the biochemical physical build of a human. This is not to deny that AI will bring substantial benefits to humanity – it already has and will continue to do so. Rather it is a matter of taking proper perspective, which then enables appropriate responses to AI.
The contention here is that the future relationship between AI and human beings will require a certain degree of discernment, understanding and wisdom. The first necessity is to ensure that AI is a net contributor to positive outcomes for human beings, society and the rest of creation. The second is to ensure new technologies are rooted in a set of fundamental ethical values. Many of these could be drawn from widely held beliefs, including human dignity, the respect and protection of life, freedom of speech, fairness in the treatment of others, non-maleficence and the rule of law. The implementation of moral values into the AI code should come through, and in some cases from, the programmers themselves. The third necessity is to promote a regulatory framework that looks at the foundational level and remains focused on the big picture. The dual challenge here is to safeguard the promotion of innovation while upholding a set of ethical standards. As will be argued, lessons from Judaeo-Christian teaching can furnish the debate with a strong moral framework.
The following study is divided into three chapters. Chapter 1 considers some of the milestones in AI history, highlighting the gradual advance from purely computational AI to AI with advanced features, such as machine learning, deep learning and pattern recognition. Chapter 2 presents two case studies. The first is an analysis of AI-led biometric identification, which showcases significant efficiency and security benefits brought to points of transit, such as national borders. The second considers the development of autonomous vehicles and how, though innovative, they are not yet deemed safe for widespread public use. Chapter 3 considers a Judaeo-Christian response to the adoption and use of AI. It emphasises the central role of humanity as image bearers of the divine, and the implications this brings for sentience as a uniquely human feature. (Terms such as ‘artificial general intelligence’, ‘superintelligent AI’ or ‘general AI’ will be used interchangeably.)
Notes to Introduction
[1] IMDB Transcript Quotes, her, https://www.imdb.com/title/tt1798709/characters/nm0424060.
[2] ‘Stanford Researcher examines Earliest Concepts of Artificial Intelligence, Robots in Ancient Myths’, Stanford News, 28 February 2019, https://news.stanford.edu/2019/02/28/ancient-myths-reveal-early-fantasies-artificial-life/.
[3] ‘AI’s Coming of Age: The Beginning of True Autonomy’, UBS Bank, https://www.ubs.com/microsites/artificial-intelligence/en/ai-coming-age.html.
[4] Jose Fumo, ‘Meet the Heroes of Deep Learning’, Towards Data Science, 13 September 2017, https://towardsdatascience.com/meet-the-heroes-of-deep-learning-648c9083ef10.
[5] Boris Babic, I. Glenn Cohen, Theodoros Evgeniou and Sara Gerke, ‘When Machine Learning goes off the Rails: A Guide to Managing the Risks’, Harvard Business Review, January–February 2021, https://hbr.org/2021/01/when-machine-learning-goes-off-the-rails.
[6] ‘Tesla’s Elon Musk: We’re “Summoning the Demon” with Artificial Intelligence’, YouTube, 24 November 2014, https://www.youtube.com/watch?v=Tzb_CSRO-0g.
[7] ‘Stephen Hawking warns Artificial Intelligence could end Mankind’, BBC News, 2 December 2014, https://www.bbc.co.uk/news/technology-30290540.
[8] ‘Oriol Vinyals: Deep Learning and Artificial General Intelligence’, Lex Fridman Podcast #306, YouTube, 26 July 2022, https://www.youtube.com/watch?v=aGBLRlLe7X8.