Conjure thoughts of just how much the person can take of alcohol, at what frequency or how fast, and for how long before he/she gets overwhelmed by a deluge of spirits—literally. Alcohol is a spirit, afterall.
One is said to be ‘under the influence’ for good reason, because that one acts possessed of spirits other than his. It shows.
If any drinker (or drunkard) would know their limits per time, then, if they do choose to drink so much, they can in some way drink optimally. That is, to drink as much as they need to not overwhelm their system either short term or long.
This is an optimization and control of drinking problem, to which many have assumed the solution of moderation as they’ve individually defined it. Some prefer to hang out with friends who won’t let them get drunk or even help them stop drinking. AA, anyone?
Now, let’s say her input is drink, and her objective is to drink as much as possible (regularly and every time), with the constraints that she ought never get drunk and for her liver to not get too full of drink to the point of growing it significantly, or breaking it. Like for several scenarios we face in life, we use model predictive control (MPC) to deal with this.
We would need a controller-actuator combo that will determine and implement the plan. The controller judges what to do based on its current perception of the state of its inner and outer world, its projection of the future using a fixed (or adaptive) model of reality, and how it chooses to determine and rate what’s best to do now, having these inputs.
finding the series of input per time (the plan)
that ‘best’ satisfy an objective (the goal)
over a horizon (the near future),
subject to known constraints (internal and external environment),
and using an uncertain model (chance and reality happening) of how the body (or world) actually responds.
It tests some series of inputs on a model to see the output it yields over several steps; then it applies the first input from the series of inputs tested that gave the best results. And it repeats this as a cycle so that the final objective is achieved.
Thus, while the objective may stay the same, the plan may be changed because of uncertainties, and chance showing up.
The game of chess is an exercise in predictive control.
You periodically evaluate your position/performance relative to your understanding of your objectives, making necessary changes as decided by the brain (controller) through your actuator. Hence one would need to own/create/evolve/develop/grow an effective model, thinking, and actuation to make any progress at all. And obviously also, we need a sensing system to monitor changes in the world and the operating environment.
How does this translate to the drinking scenario? If the goal is minimisation, the solution is trivial: don’t drink.
Drink responsibly; drink water!
MPC is the norm in automated process plants, like in breweries and distilleries. They use it to try to make optimally spirited drinks, optimally, to make as much money as they can—to drain your pocket as often as possible by your happy permission.