Goal

Our goal is to advance the intersection of mindfulness, spirituality, and technology to identify ways for artificial intelligence (AI)-based systems to support spiritual growth. While literature demonstrates the positive effects of mindfulness on self-compassion, empathy, and emotional regulation, research has largely overlooked spiritual growth despite the potential for mindfulness to foster this construct. We have also seen the scalability limitations of mindfulness interventions due to lack of qualified instructors and the emergence of technology to support mindfulness practice. However, the existing technical solutions focus on the individual instead of interpersonal connections, while simultaneously avoiding questions of spiritual growth in an attempt to appeal to a Western audience. Our idea explores how AI can expand the capabilities of mindfulness solutions to support spiritual growth and has the capacity to support projects across the discovery, development, and launch phases. Given that spirituality and mindfulness both have foundations in childhood and opportunities for promotion across the lifespan, we are particularly interested in examining how the role of AI may change across developmental stages. We seek novel strategies to this topic that include training conversational agents to recognize and answer queries that suggest a growing interest in spirituality through mindfulness practice.

Opportunity

Technologists are implementing AI-based coaches and therapists to support mental health and conversational capabilities in home health robots to provide companionship to people. Simultaneously, we see substantial investments in technology-enabled mindfulness practice. However, research and development ignores spirituality despite its known benefits, instead focusing on cognitive development and emotional regulation. AI systems are not trained to support or encourage discussions on the topic. We believe there is an enormous opportunity to discover people's interest in progression towards spirituality, develop technology to support that journey, and, in cooperation with technology companies, launch tools that improve human flourishing.

Roadblocks

Significant challenges include defining and operationalizing spirituality. Without a clear definition of spirituality it will be difficult to identify when people are starting their spiritual journey. Another challenge is outlining the spiritual development process; how do people become spiritual or engage in spiritual development? Without this understanding, it will be difficult to identify how to best support them with technology. We have no training data or framework to train an AI spiritual support system. Finally, spirituality is an abstract concept; we will need to improve conversational agents to move from answering queries to addressing more amorphous topics.

Breakthroughs Needed

How can we train conversational agents to recognize and answer queries that suggest a growing interest in spirituality through mindfulness practice? First, we need a clear definition of spirituality across the lifespan. Second, we need a better understanding of how people from different backgrounds become spiritual - what starts the journey, how they seek out information to support their growth, and the role interpersonal relationships play in the process. We need research to drive the emergence of a good metaphor (e.g., spirituality ecosystem) to help identify the role AI has in promoting spirituality. Third, we need a corpus related to answering questions about, and supporting, spiritual growth to train the AI system. Next, researchers will need to conduct a series of user studies across demographic groups and levels of spirituality to assess how well the system supports personal growth. Finally, research is needed to explore the social component of spirituality - how can an AI system evolve to support or guide social explorations of spirituality? We believe these breakthroughs are possible because they build on recent advances in science and technology, including work funded by the Templeton Foundation; the primary challenge is getting people to sustain attention on the problem.

Key Indicators of Success

By year 3, we need a framework for linking spirituality and mindfulness across the lifespan. By year 5, we will need an actionable roadmap for implementing and testing an AI system that supports spiritual growth. Failure here would indicate lack of clarity on the relationship between spirituality and mindfulness. By year 10, there should be several systems in place that are being tested across different populations. Here, we would assess whether the failure was technical in nature or a misunderstanding of the human/user interaction element of spirituality and mindfulness.

Additional Information

First and foremost, the field is lacking a high quality, accurate and comprehensive definition of spirituality that encompasses beliefs and behaviors that can be modeled. Given that there is no consensus on what spirituality looks like or feels like, it is difficult to connect research on spirituality to other holistic self-improvement practices like mindfulness. Mindfulness, defined as bringing attention to the present moment with curiosity and non-judgement, is a secular contemplative practice that is focused on promoting both personal inquiry and common humanity. If we look at the research on mindfulness, however, there is limited attention given to the spiritual development aspect of the practice despite historical links between mindfulness and spirituality. Researchers deliberately secularized the practice to make it more accessible to a Western audience. This speaks to a larger problem, where scientific and technical efforts avoid addressing questions related to spirituality or religion because of the perceived antagonism between the two practices.
Another critical area is the social component of support. Currently, developments in AI systems put heavy emphasis on helping the individual. What happens if we think about the social components of mindfulness that may also inform spirituality development? Group mindfulness practice creates opportunities to engage in informal dialog about the spiritual components of mindfulness, but how do we make sure AI-based systems support informal attempts to pursue spiritual growth or maybe even encourage it? Research on aging and spirituality indicates there may be many interactions that lead to spiritual development as well as intergenerational transmission of spiritual practices; we have to ensure that our attempts to assist human connection do not accidentally sever the connections that facilitate spiritual growth.
If there is antagonism between science and spirituality, or even a reluctance for researchers to engage in projects that actively promote spiritual development, then emerging technical solutions focused on rapport and relationship building will only emphasize superficial exchanges. A broad impact of this research program is a method for AI systems to identify when their human counterpart is looking for a deeper or meaningful conversation and to support that need. We don't know what role AI should play in the development of spirituality because we do not fully understand the process.
This is an ambitious project because we are trying to figure out how to get AI to help us pursue an abstract and ill-defined goal. The work is doubly challenging because we will have to overcome cultural barriers that have separated science and spirituality, as well as concerns that technology may be encroaching on what is seen to be a uniquely human endeavor. However, if we expect AI to become part of our daily lives, AI systems will need to learn how to help us realize our best selves, or at least not impede our efforts because systems were designed through processes that intentionally avoided difficult, but important, questions.

References:
https://doi.org/10.1037/14045-001;
https://doi.org/10.1080/17439760.2012.732102
https://doi.org/10.1080/17439760.2012.732102
https://doi.org/10.1007/s12671-014-0301-7
https://doi.org/10.1080/10447318.2019.1619259
Syracuse University collaborators include Mark Costa (Newhouse) and Merril Silverstein (HDFS). Additional collaborators include Timothy Bickmore, Robert Roeser, and Renee Kaplan.