Announcing the Formation of “Faith Perspectives on AI”
Announcing the Formation of “Faith Perspectives on AI” – a Consortium of Faith and Faith-Related Organizations Providing Perspectives on Artificial Intelligence and Ethics
Serious Risks for Human Values. Artificial Intelligence in its various forms, including machine learning, intelligent robots, and deeply nuanced predictive analytics, offers great potential for solving a wide array of problems and extending human knowledge to levels beyond existing human logic and inquiry. These solutions are unprecedented but so are the means to achieve them. Never before in human history has mankind possessed the ability to create things that can both approximate human ways of thinking and develop their own ways of thinking and communicating, ways which by definition humans may be unable to deconstruct.
The potential benefits of AI and means to achieve them must be weighed against possible risks to the most fundamental of human values: human dignity, free will, privacy, value of work, social and economic equality, and even potentially human freedom. Such fundamental reward/risk tradeoffs are not typical business questions. In a recent New York Times column headlined “Technology’s Frightful Five March On, Unimpeded”, Farhad Majoo noted that “much of what they do now, and will soon have the power to do, exceeds what we’ve ever expected from corporations.” He attributes much of that future power to AI: “They’re creating machines that could one day approximate and surpass human intelligence – a technological achievement that may come with as many complications as the advent of nuclear weapons.”
The Data Industry Response. In the past 18 months, industry-sponsored initiatives on artificial intelligence and ethics have been announced at prestigious tech-oriented universities including Stanford, Carnegie-Mellon, NYU, Harvard and MIT. In the past year, the Big Five data companies (Amazon, Google, Facebook, Microsoft and IBM) directly founded the Partnership on AI to Benefit People and Society https://www.partnershiponai.org/ The Partnership has identified seven Thematic Pillars, associated 20 non-founding partners, and very recently identified its first executive director, the former head of AI initiatives in the Obama White House. The chair of the Partnership board is the head of research for Microsoft. The key Seattle partners appear to be Microsoft, Amazon, the UW Law School’s Tech Policy Lab, and the Allen Institute for AI. Though many non-founding partners are nonprofits, only one, the Markkula Center for Applied Ethics at Santa Clara University, evidences a religious orientation. The Partnership for AI appears to be an ambitious and appropriate venture to address important questions, but lacking voices from the broader faith community.
What People of Faith Have to Offer. Questions concerning the values risked by AI are not new to the world’s Great Religions. People of faith have considered the meaning and unique value of human life for thousands of years, including, in the recent past, life-altering developments in human biology, bioengineering, pharmaceuticals, and genetic or artificial human enhancement. Yet there is little evidence that the data technology world recognizes this faith-based ethical work. Indeed, Israeli technology interpreter and futurist Yuval Harari, in his influential 2017 book Homo Deus, assumes religious groups have nothing to offer on safeguarding human values from intelligent machines and cyborgs because (among other reasons) religious practitioners have failed to keep up with technology: “[T]raditional religions offer no real alternative to liberalism. Their scriptures don’t have anything to say about genetic engineering or artificial intelligence, and most priests, rabbis and muftis don’t understand the latest breakthroughs in biology and computer science.” (278)
This ignores the reality that in the environs of Seattle hundreds if not thousands of sophisticated tech professionals at the major data companies and universities are people of faith. If mobilized, these workers and the congregations in which they worship provide a base of technical knowledge from which they and religious thought leaders can offer valuable contributions to fundamental questions about the rules of engagement between humans and machines, who will set those rules, and on what values they will be based.
Faith Perspectives on AI. Internet searches suggest that while media articles on AI and ethics abound, there is a paucity of faith-related organizations actively engaging these questions at a sophisticated level. Past interfaith experience with the Washington Global Health Alliance suggests that when different faith traditions cooperate to provide a channel for discussing shared human values with scientific organizations, the faith community voice is greatly amplified.
In that same vein, the seven organizations below have come together to explore a partnership across the faith spectrum to apply the sophisticated AI knowledge and experience of their members to ethical questions within the Thematic Pillars identified by the Partnership for AI.
- Muslim Association of Puget Sound (MAPS), Redmond
- Temple De Hirsch Sinai, Seattle
- University Presbyterian Church, Seattle
- Union Church, Seattle
- Seattle Pacific University School of Business and Economics
- University of Washington, Jackson School of International Studies, Comparative Religion Program
Faith Perspectives is currently in discussions with Seattle University to be an eighth founding partner, and with the Tech Policy Lab at the University of Washington to participate in its “Diverse Voices” policy review process. Faith Perspectives is also developing an advisory team of data and ethics experts from within the member organizations and identifying other specific opportunities to participate in the emerging debate on AI and human values.
 For example, Is AI a Threat to Christianity?, in The Atlantic Monthly, Feb. 3, 2017, https://www.theatlantic.com/technology/archive/2017/02/artificial-intelligence-christianity/515463/; Teaching Robots Right from Wrong, The Economist 1843, June/July 2017 https://www.1843magazine.com/features/teaching-robots-right-from-wrong