¹⁶While considering pain, basic emotions, and desires in a single framework is not usual in neuroscience, in recent neuroscience literature, ideas similar to the interrupt theory use the distinction between a planning system (“Model-based reinforcement learning”) and a fast system with automated reactions (“Model-free reinforcement learning”). If an agent has such two systems, a crucial question is how to divide tasks between the two systems, i.e. which one to use to respond to any particular situation (Daw et al., 2005). If fast action is required, you obviously need to use the fast system, and if there is no hurry, you can spend some time in planning; choosing which system to use is a complicated problem.. The process leading to the decision to use the fast system is then not very different from the mechanisms postulated in the interrupt theory; it has been connected to emotions by Bach and Dayan (2017).