The fish is only as big as the pond it swims in: theological perspectives on post-scaling AI

Abstract

This article examines the far-reaching implications of the Chinchilla Paper and related insights for artificial intelligence (AI) scalability, existential risk narratives, and theological reflections on AI. The Chinchilla Paper, a landmark study by DeepMind, disrupts traditional assumptions about computational power as the key to AI advancement, revealing instead that AI’s potential is fundamentally constrained by the finite availability of high-quality, human-generated training data. This insight reframes discussions on AI scalability, casting doubt on existential risk narratives that envision artificial general intelligence (AGI) as an uncontrollable force capable of catastrophic outcomes for humanity. This article underscores AI’s reliance on human-generated inputs and its inherent limitations, tempering apocalyptic fears surrounding its potential. It critiques exaggerated theological narratives that portray AI as either a catastrophic existential threat or a utopian agent of human transformation. Instead, it proposes reframing AI as a practical tool. This approach emphasizes its constraints and dependency on human oversight, promoting a balanced and pragmatic perspective on AI within both theological discussions and broader societal debates.