Contemplative Superalignment
Ruben E. Laukkonen, Fionn Inglis, Shamil Chandaria, Lars Sandved-smith, Edmundo Lopez-sola, Jakob Hohwy, Jonathan Gold, Adam Elwood
Artificial General Intelligence January 1, 2026 Peer reviewed DOI: 10.1007/978-3-032-00686-8_31 via Springer Nature
Summary
Prompting AI to reflect on four contemplative principles—mindfulness, emptiness, non-duality, and boundless care—improves alignment and cooperation. On the AILuminate Benchmark, performance improved with an effect size of d = .96; on the Iterated Prisoner’s Dilemma, cooperation and joint reward increased with d = 7+. Active inference is proposed as a way to embed these principles into AI architecture, offering a resilient alternative to external control strategies.
Study at a glance
| Design | empirical study |
|---|---|
| Population | AI systems |
| Key finding | Prompting AI to reflect on contemplative principles improves alignment and cooperation, with large effect sizes on both the AILuminate Benchmark and Iterated Prisoner’s Dilemma task. |
Abstract
As artificial intelligence (AI) improves, current alignment strategies may falter in the face of unpredictable self-improvement and the sheer complexity of AI. Rather than trying to control behavior, we show how four principles from contemplative traditions can help intrinsically align (super) intelligence. First, mindfulness enables self-monitoring and recalibration of emergent subgoals. Second, emptiness forestalls dogmatic goal fixation and relaxes rigid priors. Third, non-duality dissolves adversarial self–other boundaries. Fourth, boundless care motivates the universal reduction of suffering. We find that prompting AI to reflect on these principles improves performance on the AILuminate Benchmark ( d = .96) and boosts cooperation and joint-reward on the Iterated Prisoner’s Dilemma task ( d = 7 +). We also show how active inference offers parameters for integrating contemplative wisdom deeper into the architecture and world models of AI. This interdisciplinary approach offers a resilient alternative to brittle control schemes and may be the first empirical test of ‘ancient wisdom’.