A groove is the channel on a vinyl record where the needle goes. The needle may occasionally jump the groove when the record is scratched or dirty. Jumping the groove disrupts the flow of music and is usually experienced as jarring. The idiom “to be in the groove” means to be immersed in a particular task and thus working smoothly and efficiently. To jump the groove in this sense is also disruptive, but this time the jump is an act of will (or an output of inputs that calls itself will). - Jumping the Groove, Part I: From Dopamine to Self-Efficacy
…dopaminergic prediction errors operate over belief state representations, [which] arise when an agent has uncertainty about the hidden state of the world. - Gershman, S. J. (2017)
…exploration is hypothesized to have a dopaminergic basis...Intrinsically, the decision to explore represents a belief that neglecting more certain rewards now will produce new knowledge that may lead to greater gains later...Subjects [with greater confidence in their ability to capitalize upon later rewards, who believe their choices can determine future events] show a greater tendency to explore. - Kayser, A. S., J. M. Mitchell, et al. (2014)
Mmmm....to exploit a certain reward now or hold out for an uncertain but possibly greater reward later? If you lack confidence in your ability to achieve the later reward, you go for the immediate reward. When you're entertaining your options, you're entertaining beliefs about what is desirable, what is possible, what is within your power, and how the world works.
Beliefs serve decision-making under conditions of uncertainty. Without uncertainty, we just act. I don't "believe" the ground will stop my foot when I walk. Or that the door will open when I turn the handle. Or that air will rush in when I inhale. That's just the neural prediction and reward-seeking machinery running smoothly. It's when the machinery gets stuck that the brain shifts into belief mode to help break the logjam.
So when you hear someone say they "believe" in whatever, you know a part of them doesn't.
References:
Gershman, S. J. (2017). "Dopamine, Inference, and Uncertainty." Neural Computation doi: 10.1162/neco_a_01023
Kayser, A. S., J. M. Mitchell, et al. (2014). "Dopamine, Locus of Control, and the Exploration-Exploitation Tradeoff." Neuropsychopharmacology 40: 454-62.