Pretty much anything that can move runs on intermittent reward vs consistent consequence. Intermittent reward because sometimes in the real would things don’t work every single time, and an inconsistent negative consequence is just a high risk (and therefore more enticing) reward. Negative consequences are great short term safety solutions (one of the few things I’ll yell at a dog for is trying to run into traffic), but you need to work on teaching and rewarding a replacement behavior ASAP, because its almost impossible to create the kind of consistency needed to create an effective long-term negative consequence.

It also helps to understand extinction bursts. You know when you press a button on your computer or phone and it doesn’t work and you suddenly get the urge to hit the button harder and faster to try to get it to work? Prefect illustration of an extinction burst: when an action that typically produces a reward fails to produce that reward, the organism tries the behavior more and with more intensity. It follows naturally from the concept of intermittent reward; you know that it usually works, but that sometimes it might not, and your brain is programed to accommodate for that by encouraging you to try harder. Just remember this when you need to switch things up to curb a specific behavior; it might actually get WORSE for a short time, even if you’re ultimately doing the right thing to get the behavior to stop.