The ripples from every action, no matter how small, always affect everyone and everything in the world.
Not necessarily. Not all recursive systems are divergent. Some average out small noises. Reality is a recursive system - its next input is always the previous state.
(assuming time exists and always goes in the same direction)
I think we have definition problems here. Why assume entropy is bad? It's what nature
does. It defines the direction of time.
We living things (at least, some of us) are rare instances of localized negative entropy (in the science definition) that to exist, however, create far more positive entropy than we constitute negative (in information theory).
For example, while Bug's post is perfectly correct in info theory, the entire need for computers, the resources to make them and power to run them, is generated by us - part of the positive entropy we generate by being us - and while his scheme to reduce that positive entropy is good computer science it's still a reduction in something that
without us wouldn't even exist.
Now, if you want to get into discussion of good vs evil, and define for whom things are good or evil, that's fair game. But as beings, we are all creating more entropy than we can ever remove (unless we find something really new under the sun). Whether that is bad or not, dunno. It certainly
is.
So, I'd shift discussion to whether "random acts of kindness" are good or not. I believe that they can be - you feel better having done them - that's gotta be good. If it was a true act of kindness that actually helped someone (not just buying a drunk a drink) - that's good too, at least in my moral system. But that has nothing to do with entropy or order. Sometimes disorder is good too - if something is getting too fixated, a little revolution can be a good thing.
Sure, no one can really know the future, at least with the tools we have now. But you're going to make choices anyway - and doing nothing is just one choice of many possible ones.
The only really important question is "what do I do now". This is philosophy, not science, so much. Trying to co-mingle the two outlooks at the current state of both doesn't seem to work out too well.
In this case, scientifically, you're always increasing entropy. But philosophically, it might not be bad.