Harvard SEAS researchers have developed a detailed, generalized model for designing resonant electro-optic microcombs. They’ve shown that a single chip can host compact, programmable microcomb generators, opening paths to applications in optical communications and precision sensing.

A new study is challenging one of neuroscience’s most enduring ideas: that the brain’s reward system exists to make us feel good. Instead, researchers argue that it is built to optimize energy. Dopamine and opioids, long cast as the chemistry of pleasure, do not function as feel-good messengers but as physiological agents that optimize the body’s metabolic budget. In this view, motivation arises from rising physiological needs and reinforcement is the gain when those needs are resolved. The theory fundamentally reframes reinforcement learning. Rather than viewing reward as the pursuit of pleasurable outcomes, it proposes that learning is driven by metabolic optimization, or, the brain’s effort to minimize energetic costs and maximize gains. Within this framework, dopamine-and opioid-related processes such as habit formation, addiction, music and even social bonding are understood as expressions of a core biological principle: behaviors are reinforced when they improve the efficiency of the body’s energy regulation. In turn, dopamine-and opioid-related psychopathologies are reframed as conditions in which the brain’s energy-management system is no longer operating optimally.