Yin et al. investigated how feedback from AI-driven chatbots influences learning and brain activity. They assessed learning and knowledge retention through transfer tests and monitored brain activity using functional near-infrared spectroscopy (fNIRS). The study examined three types of feedback from the AI chatbots: Metacognitive Feedback (MF), Affective Feedback (AF), and Neutral Feedback (NF) as a control. Metacognitive feedback prompts learners to self-reflect, which has been shown to boost learning, while affective feedback aims to keep learners engaged and motivated with positive responses, thus inducing positive emotions. Both the MF and AF groups scored higher in knowledge retention compared to the NF group, with the MF group possessing higher mean scores compared to AF, despite no significant statistical difference. All groups activated the dorsolateral prefrontal cortex, a key area for memory and attention. Amongst all groups, the MF group showed the highest activity in the frontopolar area, which was positively correlated with retention scores. The study was able to provide a look into the combination of the neural and psychological aspects of learning based on feedback types, while also locating specific brain regions involved.