Dr. Yaqub Chaudhary is one of our newer Founding Experts and an important connection for AI&F with the world of Islamic scholarship in AI ethics. Dr. Chaudhary is a former/recent Research Fellow in AI, Philosophy and Theology at Cambridge Muslim College in Cambridge, England. His research interests are in the fields of AI, cognitive science and neuroscience in connection with Islamic conceptions of the mind, intelligence, human reasoning, cognition, knowledge, and the nature of perception and consciousness. In addition to professional and academic journals, he writes on current topical issues at https://www.mqbc.co.uk/ Dr. Chaudhary holds a Masters of Engineering degree and was awarded his PhD in physics by Imperial College in London. Ed.
Q: Yaqub, you have a background in electronic engineering and completed your PhD in physics. Please tell us how that led you to your current work writing and thinking about the ethical and theological implications of AI-based technologies from the point of view of Islam?
I came to work on AI from a philosophical and theological perspective through a slightly unusual route based on my interest in issues of science, technology, religion and society that I developed throughout my studies at university. In 2016, when the AI system, AlphaGo, was set to play Lee Sedol, a world champion in the game of Go, I followed the games live with great interest, and like many others, I saw this as a pivotal moment in the revival of AI, with significant philosophical and societal implications. Around a year later, a rare academic opportunity arose to do research at the intersection of science and theology at Cambridge Muslim College supported by the John Templeton Foundation. I was keen to leverage my scientific and technical knowledge to spend time to think in depth about the philosophical and theological issues accompanying the immense wave of research and development in the field of AI and adjacent disciplines.
AI needs to be considered on multiple levels. On a technological level, it is leading to new products and services that are shaping institutions, politics, society and ourselves. However, a longer-range view is that it represents the technoscientific reclamation of the most ancient questions in philosophy and theology on the nature and basis of the unique capacities of the human intellect.
It is claimed that research in AI can provide new insights into the nature of the human mind, and as such, it represents a major new site for contestation in the dialogue between science and religion. Hence, I think it is a crucial time to question and interrogate the metaphysical foundations of research in AI and the mind and brain sciences, as well as to participate in the evaluation and understanding of developments in these areas from religious perspectives, which, especially in the Islamic context, offer comprehensive metaphysical frameworks for understanding human nature and reality.
Q: In your most recent article “The Artificialization of Mind and World”, published in Zygon, you talk about the intersection of humans and machines in a shared physical and digital space through AR/VR. Could you summarize the implications for our understanding of ourselves and our experience in the world that you talk about in your article?
This article has a fairly wide scope and aims to draw together several high-level conceptual issues related to developments in AI and digital technologies. The three key issues are world, mind and knowledge.
First, in relation to the world, I discuss the implications of the digitalization of space and time, and how increasing portions of our world are being recast into artificial digital microworlds.
Second, I highlight that artificial agents exist against the background of these artificial worlds even when they appear to act in the physical world, so interacting with an AI agent is interacting within the envelope of its world, even as it learns finer representations of the aspects of the physical world it is designed to operate in. The analogue world we inhabit needs to be digitally ‘reformatted’ as digital data for input into an AI or ML system, including our own human nature, which is hollowed out leaving only what can be peeled off as quantifiable data such as biometrics, psychometrics, behavioral patterns, preferences, purchase history, and so on.
Finally, I discuss how science itself is becoming increasingly artificial as what comes to stand as knowledge and understanding is increasingly mediated through these artificial constructs of mind (intelligent learning systems) and world. The key point is that the objects of analysis have been abstracted, severed even, from the real world and transformed and delimited by only those attributes that are quantifiable and digitizable.
The theological significance of these three issues are first, the re-emergence of an immaterial space, similar to an unseen sacred realm, that is coexistent with physical space in the modern worldview; second, new conceptualizations of human nature; and finally, veneration of the machine gaze, whereby concepts of divine foreknowledge and providence become ascribed to AI as we cede judgment and control of society to AI.
A further problem occurs when the aforementioned artificial worlds are re-mapped onto the real world such that nature is seen through the veil of artificiality. In some sciences this veil is so opaque that nature is not visible at all and it is not nature that is subject to observation and analysis but artificial nature. AI and machine learning have led to an intensification of the computational turn in research in the digital era which “reframes key questions about the constitution of knowledge, the processes of research, how we should engage with information, and the nature and categorization of reality,” according to boyd and Crawford writing in 2011. Now, new AI techniques and high-performance computation, amount to new “digital optics” that aid scientists to uncover new insights in the vast quantities of data now available. A similar argument has recently been made in an online article (with an excellent accompanying illustration), which describes these AI-enabled digital optics as a “nooscope” – “an instrument to see and navigate the space of knowledge.”
The Zygon article also discusses how the above three conceptual issues relate to three major transformations that we are currently in the midst of. First, anthropocentrism is being succeeded by mechanocentrism; second, the totalization of a digital/informational ontology is altering our conception and understanding of reality; and finally, a new tension has arisen between dualism and monism.
Q: On balance do you see more opportunity or more risk for human dignity and flourishing from advancements in artificial intelligence and the applications it powers?
The prevailing view in technology studies is the social shaping of technology, so the opportunity for human dignity and flourishing exists only in so far as we know what these terms mean and in so far as we make them part of our key objectives in new applications of technology.
However, the evidence from the past century of technological innovation would seem to show that we don’t really know what human dignity or flourishing means anymore and rather than being key objectives, AI intensifies ambitions of prediction, manipulation and control of nature and human society.
We know that the development of AI is strongly implicated in the unscrupulous world of the data brokerage industry that is collating vast quantities of information that is being used to manipulate people into purchases, deny credit applications, raise medical insurance rates of people who are already suffering illnesses and so on. Recently in the UK we saw how the injudicious use of algorithms disabused an entire cohort of students from the fruits of their studies by downgrading exam results based on preexisting social inequalities.
In this connection, a recent paper has argued how machine learning is being introduced into the educational sector in a way that tends towards a new form of “behavioural governance” that works against, rather than in support of, student autonomy. Furthermore, interacting with AI agents in the school and home environment may create “…a lifelong intuition in children that software has feelings that can be hurt, that it’s an intelligent being to be respected–or even an authority to be obeyed,” (Mike Elgan).
Another recent paper has discussed the colonisation of the home by smart devices, in which the author argues that rather than augmenting domestic life for the benefit of family members, these AI-based smart devices represent a new site for platform capitalism that intervenes in and pacifies family life into an “engineered simulacra of domesticity” that serves as a vehicle for domestic consumption. Many others have written on similar themes and all of this seems to me to be an affront against human dignity and flourishing, rather than in its favour. To be clear, my position is not against digital technology or AI. What I oppose is how corporations are forcefully insinuating themselves through AI as third parties in every interaction and area of society for manipulative and extractive purposes.
Q: You speak and write for popular audiences alongside your academic work. How can the faith world better engage these opportunities and risks, starting with your own faith perspective as a Muslim?
Islam has immense intellectual riches to be drawn from in addressing the challenges related to AI raised above. For my purposes, it offers a firm standpoint, that is, a comprehensive worldview and metaphysical framework, through which to engage a topic such as AI with immense conceptual complexity that defies disciplinary boundaries.
I think that those engaging AI from a religious perspective should be weary of their discourse merely conforming along predetermined tracks or being thrown off the tracks of meaningful discourse altogether by having to attend to pseudoscientific distractions such as consciousness uploading, merging with AI and technological singularity. Some discussion is required in these areas, however, many of these issues dissolve away on deeper inspection of more fundamental aspects of AI technology.
One problem is that as AI has come to public and scholarly attention, certain discursive formations have quickly been adopted by its proponents in industry or academia, and slavishly followed by others who lack the technical know-how to disentangle the myths and magical thinking about AI. This, for example, has led to simplistic discussions on “AI ethics” and the consequent spectacle of “ethics theatre” – the mere pretense of adhering to and operating according to ethical principles.
My view is that it is the religious standpoint that offers the potential for the deepest possible examination of AI and for understanding its potential for enhancing or diminishing human dignity and flourishing. It is the world’s religions that have the best record of teaching the most balanced use of the powers of the human intellect. I think that followers of religion need to generate their own discourses on AI and demonstrate the value of their thought leadership.
Religious discourse on AI needs to begin with asking the relevant questions and undertaking the right kind of inquiry into the presuppositions embedded in standard discourses as well as developing an in-depth understanding of issues such as the nature of the technology, its history and context in which it originated and developed, and with what intentions and ideological background.
Q: Where do you see your work taking you next?
There are now dozens of major institutions addressing issues of AI and society in a secular context. However, given the role religion still plays in society and in people’s lives, and its rich resources for enhancing the condition of humanity, society and nature, I think it is necessary to develop the capacity to address issues from within the scope of religious traditions themselves. My work at the moment is an effort in this direction, that is, to establish some preliminary intellectual groundwork for deeper examination of the issues arising from AI and digital technologies, as well as at the frontiers of science more generally. In the future, I hope it will be possible to carry this research forward with larger teams of researchers and collaborators.
In the meantime, I will be speaking next week, on 16 September, on some of the topics raised above as part of a conference on Artificial Intelligence and Religion organised by the Center for Religious Studies at Fondazione Bruno Kessler (FBK). The complete programme for the fortnightly series with a superb range of speakers is available here and is likely to be of interest to this community.