Diverse Approaches to Applying Ethics to AI

Global Perspectives on AI Ethics Panel #12

AI Ethics: Global Perspectives is a free, online course jointly offered by The Governance Lab (The GovLab) at the NYU Tandon School of Engineering, the Global AI Ethics Consortium (GAIEC), Center for Responsible AI @ NYU (R/AI), and the TUM Institute for Ethics in Artificial Intelligence (IEAI). It conveys the breadth and depth of the ongoing interdisciplinary conversation around AI ethics. The course brings together diverse perspectives from the field of ethical AI, to raise awareness and help institutions work towards more responsible use.

The twelfth installment of Global Perspectives on AI Ethics was held on Friday, February 17th. This month, we were joined by the following distinguished faculty for an exciting discussion around the many different approaches to AI ethics:

• Amana Raquib, Assistant Professor of Religion and Philosophy at the Institute of Business Administration Karachi (Karachi, Pakistan)
• Marta Galceran, Research Fellow — Global Cities Programme at the Barcelona Centre for International Affairs (CIDOB) (Barcelona, Spain)
• Viviana Polisena, Professor-Researcher at the Catholic University of Córdoba (Córdoba, Argentina)
• Yip Fai Tse, Researcher at Princeton University’s University Center for Human Values (Princeton, NJ)

The discussion was co-moderated by two of our course leads: Julia Stoyanovich, Director of the Center for Responsible AI at NYU and Stefaan Verhulst, Co-Founder and Chief Research and Development Officer of The Governance Lab. Over the course of the conversation, the panelists explored a wide range of topics from the value of a virtues-based approach to ethical AI to the role of AI in cities and the need for new governance models around AI technologies to the inclusion of animals in the current discourse.

A virtues-based approach to ethical technologies

To open the panel, Julia invited Amana Raquib to open the panel, focusing on her module: The Islamic Philosophy of Techno-Social Good, and the role of philosophy in guiding the impacts of AI and other emerging technologies on human life.

Amana began by pointing out that a virtues-based approach can help guide society’s use of technology to promote social good by taking into account a more holistic understanding of the impacts of different technologies on human life. She compared a virtue-based approach to a consequentialist approach citing the somewhat more short-sighted and narrow perspective of a consequentialist approach in predicting the repercussions of technology. In talking about a virtues-based approach, Amana said:

“Instead of looking at individual technologies and their specific consequences, we need to have a holistic assessment of how it affects individuals and communities, not only in terms of their material and physical well-being, but also their moral, spiritual upliftment and how it actually serves those broader and greater objectives of preservation of human life.”

Building upon the idea of a virtues-based approach, Julia asked Amana about the different ways in which we can cultivate diverse religious and secular virtues to promote the ethical use of AI. Amana highlighted the importance of drawing on shared understandings of a good life and human good as foundations for these conversations. She cited a recent paper she wrote where she concludes that the universality of shared virtues offers a key intersection for different religious and non-religious cultures. Moreover, she argued that diversity and inclusion are also important during the design and development of new technologies to ensure that technology is designed “for humans” rather than “to or at humans”.

When asked to provide an example of this approach in action, Amana shared the case of AI chatbots being used to ease loneliness. From a technological point of view, the discourse focuses on the question of what these chatbots can accomplish and how, looking at their ability to emulate human speech and engage with society. From a social and religious perspective, however, the more profound question to ask would be why are people feeling this loneliness? Amana argued that it is these more human questions which ought to be driving the development of emerging technologies

Before turning the floor over to Marta for her opening remarks, Stefaan asked Amana to elaborate on potential actions we could take to steer the development of these technologies towards answering the deeper, more profound questions at hand. In response, Amana emphasized the role of shared guiding principles, such as compassion and community-building, in determining the design and impact of new technologies. She concluded her remarks with a critique of the rapid commercialization of technology, which she feels limits the amount of time and resources directed towards understanding the social and ethical implications of new technologies.

The challenges and opportunities of urban ethical AI

Following this conversation, Stefaan turned the stage over to Marta Galceran for her opening remarks on the role of ethical urban AI in building trustworthy cities.

Drawing on her extensive research experience in city diplomacy and urban governance, Marta first pointed out the huge potential for AI in improving service provision and public administration in cities through automated decision-making, innovative technology-driven systems and expanded service options. Unfortunately, along with this potential, Marta identified 4 key challenges facing ethical urban AI today.

The first two challenges, bias and a lack of regulation in the ecosystem, Marta argued, are interrelated, saying:

“We cannot think about the deployment of [AI] technologies in cities without thinking about the regulation of these technologies. We need to establish the right procedures, the right normative frameworks, to safeguard fundamental rights and digital rights in cities…some cities are [deploying AI technologies] in a vacuum because we are in a context of fragmented AI regulation.”

In spite of this fragmentation, Marta was optimistic that the lack of regulation would not persist given the number of cities around the world drafting and implementing their own regulatory frameworks to govern the use of AI. This bottoms-up approach in cities, she felt, was more effective as the impacts of AI technologies at the city-level are oftentimes more evident, and therefore are easier to adapt and respond to.

Moving to the third challenge, Marta highlighted the gap when it comes to transforming principles into actionable practices. A large part of this challenge, she found, emerges from the more abstract nature of principles and the absence of concrete definitions to help guide actions. In terms of bridging this gap, Marta felt that innovative and collaborative approaches to effectively match principles with practices would help drive the adoption of ethical urban AI, citing the work of the Cities Coalition for Digital Rights as one example of this.

Last but not least, Marta called attention to the limited capacity in the public sector when it comes to working with emerging technologies. A lack of in-house technological expertise often leaves public sector actors dependent on external, private sector contractors, creating a gap in their collective understanding of the ethical implications of an emerging technology. The solution Marta offered includes transparent public procurement systems and better safeguards of individual rights.

As a follow up question, Stefaan asked Marta about the adoption of ethical urban AI in other parts of the world pointing out that the majority of her examples were focused on cities in Europe and North America. Marta responded by calling out the critical lack of representation in the current discourse around cities and AI of stakeholders from the so-called Global South. In the Cities Coalition for Digital Rights, for example, at least 60% of the cities represented were from Europe. Marta shared that barriers to access are limiting efforts to be more inclusive, however new partnerships with global actors, like UN-Habitat, can help bridge these gaps by creating new avenues for knowledge sharing.

The intersection of humans, nature and technology

Next, Julia invited Viviana Polisena to discuss her module, Global Pact: Birth and Organization of Territories. Viviana’s remarks were helpfully translated by Marta.

To open the conversation, Julia asked Viviana about the role of the phenomenological paradigm in establishing a global pact to govern the use of emerging technologies. Viviana began by explaining the intricacies of the phenomenological paradigm which, she shared, proposes a combined ethical, scientific, metaphysical and theological solution to the challenges currently facing the natural world. The paradigm calls for a reconception of the natural world and our place within it. Her central argument was that by understanding how humans, nature and technology are related, we can achieve a new global pact to address the negative impacts of technology on the natural world.

Building on this idea of the impact of technology on the natural world, Viviana argued that the rapid advancement of technology disrupted an existing balance between the natural world and the human world, wherein the two worlds were kept separate from one another. Given the pervasive nature of technology today, however, Viviana asserted that this separation no longer exists. Instead, we must now consider the impacts of technology in a more holistic manner, taking into account these two worlds in tandem in a kind of “digital expanded reality”.

Diving deeper into the concept of a global pact as one solution to the challenge at hand, Julia asked Viviana about some of the potential benefits of a global pact in driving the ethical and sustainable use of technology. Before outlining these benefits, Viviana mentioned that we would need to approach a global pact from a more holistic perspective under the new paradigm she proposed earlier.

The three benefits Viviana proposed were: the ability to self regulate and to regulate human creations more effectively; the power to overcome differences and divisions; and the creation of new, more inclusive territories. Viviana further expanded on these benefits later on in the panel.

The case of AI ethics and animal rights

Finally, Stefaan opened the stage to Yip Fai Tse to talk about his work on the case of animals when it comes to understanding and evaluating the impact of AI and other emerging technologies.

Fai explained that his research work focuses on studying the existing AI ethics discourse to understand its treatment of non-human animals. In his work so far, he has found that virtually none of the current principles, policy frameworks or academic research around ethical AI protect or even account for non-human animals, even though AI has dramatic impacts on non-human animals. In explaining the current gap in the discourse, Fai said:

“Our research finds that AI does impact non-human animals. If there are impacts on non-human animals, especially non-human animals that are proven to be sentient and have the capability to have experience such as pain or pleasure or interests, then this kind of impact on them matters. Ethically speaking, it is imperative, or it is an obligation, that we identify these impacts, and then try to reduce the harm, or maybe even bring benefits to them.”

Citing the examples of self-driving cars, drones and factory farms, Fai highlighted the important question of humaneness in the treatment of animals by these technologies. When it comes to integrating the case of animals into the mainstream discourse and practices, Fai advised we adopt a broad and general perspective, starting from the simple idea that AI impacts non-human animals in many ways. He shared that this often comes as a surprise to stakeholders in the ecosystem who may not have otherwise realized that AI technologies do impact non-human animals. Furthermore, he advocated for the increased awareness of the impact of technology on non-human animals not only in the realm of academic research and technological innovation, but also in the policy sphere by updating AI ethics principles and expanding our moral circle to be more inclusive.

Shifting gears slightly, Stefaan asked Fai about the possibility of expanding the AI ethics discourse further to account for the environment and planetary health more broadly. As a utilitarian, Fai approached this question through the lens of utility, or whether or not an action is producing more harm than good or vice versa. From this perspective, he felt that we would generally prioritize sentient beings over non-sentient beings and that if we were to account for non-sentient beings, we would favor those non-sentient beings we depend on for survival. That being said, Fai did agree that a broader and more inclusive approach to ethical AI may be helpful, with the caveat that it might fail to account for smaller, more granular issues on the individual-level creating new ethical dilemmas.

Moving forward: the value of inclusion and an optimistic outlook

Following an engaging question and answer session, where audience members and the panelists unpacked a number of thought provoking questions and shared deep insights, ranging from the need to balance religious and secular perspectives to moral uncertainty and agency to the role of city planning in achieving well-being, Julia and Stefaan asked each of the panelists to share their final reflections.

Fai concluded his remarks with a final call for the inclusion of non-human animals in the AI ethics discourse, pointing out that since non-human animals cannot engage in the discourse themselves, the responsibility to defend their rights and their well-being falls entirely on the shoulders of humans.

Marta addressed the audience question regarding the importance of regulation in the AI ecosystem, once again emphasizing the value of a bottom up approach to regulation, especially in terms of promoting a more inclusive and diverse discourse.

Viviana advocated for the value of inclusivity and collaborative action in achieving sustainable change and helping to mitigate the impacts of technology on humans and on nature.

Finally, Amana had the last word, offering an optimistic outlook on the future of AI ethics, where open conversations and the sharing of insights across diverse sectors and contexts could help effect positive change in the long run.

To watch this webinar and others, visit our course site here. We post new modules and other exciting content every month. To receive updates on the course, please sign up at http://bit.ly/ai-ethics-course.