This is the only sane answer I read here so far. I'm surprised that so many people still believe that depression is at large some kind of pathological condition

This is the only sane answer I read here so far. I'm surprised that so many people still believe that depression is at large some kind of pathological condition. I'm even more surprised when people accept the most addictive drugs in existence as a "treatment" for anything. You make some really great observations in your post. I found them very enlightening, eye-openers. However, I have to disagree on the fact that depression is usually/mostly a change of your perceptual model. As you mentioned, depression is the state during which your prediction of the future (correctly or incorrectly) is that any action you take will result in a highly unpredictable, and mostly dangerous/damaging/negative feedback. Bad enough feedback for the best action plan to be doing nothing at all. There are two main, and very important types of depression branching off here. One is the type where your prediction is wrong. The other is the type that your prediction is right. There are many cases where your modeling of reality is now flowed or outdated. Maybe you grew up with negative or mentally ill people, and you know no better. Maybe your school teachers were assholes. Etc. In this case someone needs to show you, possibly through demonstrations, that your perception of reality is actually flawed. Off the top of my head, most psychotherapy, like cognitive behavioral therapy, offers just that. Your therapist will try to prove your model of reality to be wrong (even if it's not, but that's another discussion), and convince you that things are better, to look at things positively etc. Some times, for some people, that works. Other might refuse to update their model and choose to ignore reality. Others find dysfunctional ways to cope with life. Others will get a terrible therapist with a worse model than theirs. Many things are possible. The other main category of depression is the type where your model of reality and your prediction is correct. Like a dog or lab rat that gets zapped with 50% probability no matter what it does. At first it will try really hard to figure out what's going on, and in the end it will figure that trying isn't worth it and it will stop. You can give that animal all the ketamine in the world, and it still won't be happy, because their environment is flawed, not their perception or learning abilities. In real life, examples include dealing with unstable people (bipolar, narcissistic etc), being stuck at an actual dead-end situation with near-zero chances of recovery (homelessness, bad chronic illnesses), etc In this case, someone needs to fix your environment (if possible, ie homeless or dealing with unstable people) and not your perception.