The logical road to human-level AI is a dead end. Human-level AI requires compassionate intelligence to guide interaction and build applications and operating systems of the future. The cognition of such an agent includes Metacognition: thinking about thinking, thinking about feeling, and thinking about others' thoughts and feelings. We summarize the core meta-architectures and meta-processes of EM-2, a meta-cognitive agent that uses affective inference and an irrational TMS, the I-TMS, to resolve the turf war between thoughts and feelings based on agent personality and social and cultural knowledge as well as logic. The work is inspired as much by a human mind training process from India as it is by the 17th century commonsense philosopher David Hume. Although separated by more than 2000 years they both present a philosophy of mind where emotion is the antecedent to logical thought.