We promise unparalleled technological advances for today’s most pressing problems. At the same time, I am aware that biases in AI can harm humans, and that humans may unknowingly internalize those biases as fact even after they stop using the algorithm. we are aware of. Whether AI has the potential to do good or harm depends on eradicating biased data and the human and institutional biases embedded in those datasets. I am.
One critically undervalued and misrepresented data set is the importance of religion and spirituality in the lives of people of faith, and most humans.
Most of humanity (84%) belongs to a religion. This proportion is expected to increase over time. Meanwhile, an AI-powered study of more than 30 million documents found that 63% of faith-related digital content deals with faith negatively. Additionally, 11% of digital content that mentions faith is negative in an extreme way, including hate speech, and a total of 74% of all faith-related digital content is negative to some degree. According to multiple global surveys, the majority of people say that expressions of their faith are stereotypical, sensational, negative, or simply non-existent.
As the digital environment provides training data for AI, these archetypes of negative beliefs can have a negative impact on human flourishing. Religion and spirituality are linked to better well-being, mental health, civic engagement, prosocial behavior, and longevity. Religious stereotypes can fuel violence against marginalized groups. Training datasets filled with negative and even hateful religious expressions risk exacerbating violence and reducing the positive impact that is sorely needed on individual well-being and society.
If we want AI to be truly inclusive and able to help humanity thrive, training datasets must be audited to prevent unfair and inaccurate biases against religion and spirituality.
Experts and stakeholders actively advocate for accurate and representative faith narratives in our collective cultural narrative to world leaders, the media, the public and private sectors, and other influencers. There is an urgent need to apply pressure to do so. By encouraging representative stories from all faith and spirituality types and preventing bias and discrimination, we can train AI systems with better datasets.
In the face of rapidly accelerating and often incomprehensible technological change, AI has the potential to accelerate us toward devastating harm or unimaginable prosperity. The power of AI to enhance the beauty that makes us uniquely human depends on whether what powers it expresses the noblest and most transcendent human virtues.
Spirituality has nothing to do with the impact of AI on human flourishing. It’s essential.
