Artificial intelligence is reshaping relationships by providing conversation and companionship, and reshaping how people work. For children, it is making toys interactive and data-driven, and it is mechanizing and perhaps dehumanizing healthcare. The speed and magnitude of these transformations is breathtaking.
Effective leadership of the development, deployment and monitoring of AI requires addressing risks to people and the environment. It also calls for ensuring equity – fair access to AI’s benefits and fair mitigation of potential harm. Pope Leo XIV set this tone in his recent encyclical, Magnifica Humanitas, about the human promises and perils of AI.
We are scholars and leaders in human ecology – the interdisciplinary study of relationships among people, families, communities, society and the natural and built environment. We ask a central question: How can AI be truly human-centered, so that communities – and all forms of life – can thrive?
We think the key is having all stakeholders – families, educators, engineers, designers, policymakers and citizens – involved in designing, coding, testing and monitoring AI. Innovation can come from not just programmers but also ethicists, policymakers and AI users. It can also come from experts who study implications for human behaviors and systems, with inclusion of historically marginalized people’s perspective, ideas and talents. Co-creation also makes it easy to “crash-test” innovations for harm before wide product release and for a balance between productivity and safety.
Pharmaceutical development offers one lesson. The development of lifesaving medications can be thwarted by too much caution in not tolerating some potential side effects. But lack of guardrails can result in drugs that have deadly side effects or simply don’t work. In many industrialized nations, evidence-based regulation has led to a balance between making drugs available and keeping people safe.
Although AI is not currently meaningfully regulated, similar kinds of guidelines could optimize benefits and prevent harm.
Principles for humanistic AI
Practical ethics and policy guidance are key to ensuring that the coming AI world is safe, equitable and environmentally responsible. Human ecological principles suggest that the best way to proceed is to analyze every AI proposal, product or service in terms of whether it supports human well-being and the global ecosystem. That includes supporting connectedness among individuals, in households, workplaces, communities and society.
To those ends, we offer eight principles as the beginning of a conversation among all people who may be involved in AI’s development and use.
-
Begin any consideration of AI development or application with a human ecology frame, by asking, “How can humans use AI to promote thriving for all?” This perspective is broader than the primarily technological question of “What can AI do?”
-
Consider ethical frameworks and guardrails to prevent and mitigate harm. If governments, universities and the private and nonprofit sectors include accountability in the design, planning and implementation of AI, then people are more likely to adopt the new technology than to mistrust it. Furthermore, if people consider the impacts on relationships among humans, institutions and the Earth, then the public can properly weigh the benefits against the downsides. An important way of preventing harm is by crash-testing best ideas about AI applications and openly imagining the worst things that can happen.
-
Study where AI use may be appropriate and not. If AI co-creators consider scientific evidence about human development, from fetus to elderhood, and recognize the inherently different needs at different phases of life, then AI stands a better chance of strengthening, not substituting for, human care.
-
Developers can pursue AI designs that respect and support the social habitats that sustain human thriving. For example, recognizing that a well-designed AI model can briefly engage a young child while a parent may be preparing dinner is far more realistic, and in our view appropriate, than claims of AI replacing interaction with human caregivers. Well-crafted AI wouldn’t supplant the caregiver-child connection but would help preserve the health of relationships, reduce strain on parents and allow for better connection.
-
Educators can develop strategies at all levels, from K-12 through college, in which teachers and students use AI to help accomplish learning objectives and develop mentoring and nurturing relationships. Ecosystems for learning thrive on relationships among teachers, peers and families – not algorithms. These relationships foster critical thinking, perspective, creativity and problem solving. One way to improve human well-being is to avoid the harms of digital dependence, including physical inactivity and social isolation.
-
Rather than hew to the well-known mantra of “move fast and break things,” AI developers can implemented innovations at a pace and scale that respect the risks of moving too quickly. Chaos is more likely with sudden disruptions to social equillibrium.
-
Policymakers and planners can foster a civic AI culture that serves the common good. Equity is a critical measure of healthy social systems and so, we believe, should be a primary goal of AI development. One way to accomplish that is developing, deploying and monitoring AI in accordance with concepts such as the triple bottom line, which recognize not only profit but people and the planet.
-
Policymakers can also develop metrics to assess the impact AI has on many aspects of human and environmental well-being. Every digital action has an ecological shadow. An accurate accounting of AI’s costs and benefits can help establish just and sustainable deployment.
Guiding AI development
Like any ecosystem, the digital dimension of our lives is evolving, but people can take an active role in guiding that evolution. AI will be what we make it. A holistic approach can ensure that AI strengthens the webs of life, learning and care that make us human.
