Abstract: In recent years, a kind of renaissance of reinforcement learning applications has occurred separately in two domains. The first is the deployment of reinforcement learning systems to learning problems that seem on their face to be extremely complicated, and thus amaze our sensibilities about the kinds of problems solvable by such ‘simple’ learning systems. The second is the use of the ‘reinforcement learning platform’ to inform and interpret neurobiological experiments that seek to connect reinforcement learning models to various features of neuromodulatory signaling. In rodents, these latter applications abound and now make use of a collection of superb molecular tools for fast, selective detection of important neuromodulatory transmitters or selective stimulation of their parent neurons. This rodent work includes fast measurements of dopamine, noradrenaline, acetylcholine and more. There has been no comparable work in humans, which opens up a gap for understanding the neuromodulatory substrates of human decision-making. In this talk, Dr. Montague will present new methodology for sub-second detection of monoamines in conscious humans and demonstrate its deployment in a range of simple valence and arousal processing tasks.