Active open mindedness and mindfulness – Vervaeke (www.youtube.com)

Cognitive scientist John Vervaeke warns that as AI takes on more decision-making, human cognitive agency risks atrophy unless we deliberately practice two opposing mental skills: structured inference (reasoning within a frame) and insight-driven reframing. He illustrates the difference with the lily-pad puzzle — where stepping through inference leads many to the wrong answer (day 10) while an insight (recognizing exponential doubling) yields the correct answer (day 19). To stay adaptive you need active open-mindedness — disciplined, bias-checking reflection that constrains premature jumping — and mindfulness — practices that quiet inferential machinery so true insights (frame shifts) can occur. Crucially, these are opponent processes: each must be available but neither maximized in isolation. For the AI/ML community this is a practical prescription for preserving human oversight and better human–AI collaboration. Designers, researchers, and operators should value workflows and interfaces that both slow and enable cognition: tools that encourage explicit hypothesis testing and bias discovery (supporting active open-mindedness), and moments or modes that permit re-framing and creative leaps (supporting insight). The balance helps mitigate automation bias, improves model auditability and decision alignment, and suggests training programs that cultivate reflexive reasoning and mindful reframing as part of responsible deployment and governance.