Means and Ends for AI and for Faith

Last week my desktop computer suddenly stopped booting up.  I looked at replacement options while awaiting the technician’s diagnosis.  The friendly sales rep’s first question was “What will you use it for?” When I told her about my work with AI and Faith, she asked politely but dubiously, “does religion say anything about artificial intelligence?”

That is our essential question, isn’t it?  AI and faith seem like opposites.  Indeed, twenty months ago, our early Founding Members decided on this name to play off a sense of unlikely bed fellows.  We set out to demonstrate that, in fact, these two ways of thinking pair well in the quest for human flourishing.

A lot has happened since we launched.  We have grown our community of Founding Members to nearly fifty experts around the country.  We are beginning to replicate our local collection of faith congregations and academic programs in other key tech centers.  We are highlighting the work of our Founding Members, institutional partners, and other people of faith on our website and in our monthly e-Newsletter; by participation in conferences and programs; and through personal interaction in the AI ethics community.  And we are finding…

Pairing AI and faith works.

This is evident in the content of articles and book reviews regularly contributed over time to our website and newsletter by our Founding Members and in the extensive body of writing by  Founding Members who have just joined us like Pastor Paul Taylor in Palo Alto, data scientist Elias Kruger in Atlanta, and Middlebury College computer science professor Matthew Dickerson, author of the influential book The Mind and the Machine.

It is also evident in the report in this issue of the Newsletter by our Founding Member Brian Green from last week’s Vatican Conference on Common Good in the Digital Age.

Influential outside voices are adding further support.  In Tools and Weapons: The Promise and the Peril of the Digital Age, published last week and reviewed in this issue of the Newsletter,  Microsoft President Brad Smith and his co-author Carol Ann Brown call for “seats at the [AI and ethics] table not only for technologists, governments, NGOs, and educators, but for philosophers and representatives of the world’s many religions.”

And at this past Wednesday’s downtown Seattle Rotary Club meeting, Canadian Consul General Brandon Lee, singled out faith leaders as important contributors to establishing guard rails for AI.

Overlapping Ends

The relevance of the faith world to the creation of guard rails for AI should not surprise since  faith and AI operate in much the same areas of concern and even share some common ends and means, though they differ fundamentally in approach and practice.

As to ends, both AI and faith sectors fundamentally seek to better understand human behavior and the world around us.  AI promises to identify never before seen patterns, correlations and insights in human health, material science, the laws of biology, physics, and everything in between.  The world of faith (including many AI professionals who are people of faith) weighs such revelations against long-developed belief systems concerning human origin, human nature, good and evil, justice, and the elemental forces of the universe writ large.  In a sense, AI’s deep dive into machine learning and pattern recognition has moved it closer to a “leap of faith” of its own: no longer requiring strict proof of causation in the classic scientific method for some in AI, makes high degrees of correlation based on a large enough data set sufficient to move scientific understanding forward.

There are also similarities regarding “ends” in the sense of potential “outcomes.”  The faith world anticipates outcomes ranging from perfect fulfillment to total apocalypse (and sometimes both).  So do contemporary discussions of AI outcomes.  Compare Ray Kurzweil’s The Singularity is Near: When Humans Transcend Biology  with James Barrat’s Our Final Solution: Artificial Intelligence and the End of the Human Era.  Brian Reese contends in The Fourth Age: Smart Robots, Conscious Computers and the Future of Humanity that which way you come out on these predicted outcomes turns on your beliefs about the nature of the universe, of humans, and of self.   As our reviewer noted, these beliefs are fundamental questions of faith as well as science.

Overlapping Means

The worlds of AI and faith arguably also work through similar means.

Eugene Peterson, the noted Bible paraphraser, asserts in Living the Message that the two essentials of human fullness are: “Intimacy: we want to experience human love and trust and joy.  Transcendancy: we want to experience divine love and trust and joy.”  (244-45)

Big Tech offers one version of intimacy to users of social media and interpersonal applications and a different version to the end-purchasers of its data and AI-powered products. Playing off the fundamental human desire to know and be intimately known, Big Tech encourages people to share personal details and images on social media and through consumer loyalty programs with few or no boundaries.  AI companies then collect this data; derive from it individual interests; and package it as deeply granular knowledge that businesses, politicians, and researchers can use to sell things, influence choices, and derive even better means of manipulation.

Intimacy in the faith world by contrast operates best in real interpersonal relationships that are thoughtfully boundaried by teachings, mores, and beliefs about love, justice, marriage, community, character and moral obligation.  This happens in real places: religious gatherings, schools, small groups of people gathering in homes, social gatherings of all sorts.  The opportunity for genuine intimacy through real social interaction is fundamental to these gatherings.

Given human nature, intimate relationships in the faith world admittedly often go awry. Pederastry, terrorism, extremism, and fear of the other have especially dominated recent headlines about religion.  But the AI world of unboundaried relationships is even worse.  Consider the documented PTSD suffered by Facebook’s huge cadre of contract censors who are tasked with applying ever evolving guidelines to the worst that man can do and share. Faith systems know that boundaries are needed up front, not after the fact.  And they know that outside of a model that provides a basis for loving interaction and hopeful outcomes, human intimacy is more likely to produce a deep sense of fear and meaninglessness than Peterson’s “trust and joy”.

Turning to transcendence, in this Sunday’s New York Times opinion section, University of Washington historian Margaret O’Mara calls Big Tech’s transcendent credo The Church of Techno Optimism:  “a decades-old belief system, supported by lawmakers of both parties, which holds that networked computers are tools of liberation (even if it’s not entirely clear who is being liberated from what) and that more connection, more transparency and more powerful technology will somehow ‘make the world a better place’ (even if it’s not entirely clear what is better or for whom).”

AI’s version of transcendence often seems to simply substitute man as creator and controller.  Bruce Baker’s post “Playing in the Cybernetic Meadow” describes the attractions and possibilities of coders becoming homo faber – coder as maker – by programming semi-autonomous machines to do their bidding.  The drive to recreate human functionality in a more controllable form as mechanical agents or chat bots seems especially rooted in grasping for transcendent power and control.

By contrast, faith systems that have stood the test of time look beyond man and man’s own creations toward a deity who is sovereign, just, loving and trustworthy.

Scope and Scale are Huge

Finally, it is worth noting a common scope and scale for AI and the faith world.  They both transcend geopolitical, economic, and social boundaries.  Like the subjects of AI-powered inquiry, faith systems’ subject matter ranges from the cosmological origins of the universe to the depths of human neural activity, and even further, beyond, into a non-material world.  As to scale, assuming the truth of the oft-cited statistic that three-fourths of the world’s seven-plus billion people follow a religious tradition, it seems likely that roughly as many daily prayers of worship and petition are served up to divine beings as search requests to Google – about 5 billion daily according to Internet Live Statistics.

In sum . . .

There is enough overlap between the ends and means of faith systems and Big Technology to expect that each has something valuable to say to the other.  Thus, AI and Faith’s end goal:  to promote good and guard against evil and promote  human flourishing through the union of deep technical knowledge with systems of belief and underlying values transformed into wisdom by long inquiry, practice, and revelation.  And its means:  to bring together in community sophisticated technologists with ethicists, theologians and philosophers from faith systems, and share the lessons learned from this engagement across the broad world of faith institutions.  We’re eager to get on with it!

X