I can vouch for the value of this approach. My apartment complex has an informal back path (paved by generations of feet) that people use to get to the nearby university. It also gets some occasional use from the local unhoused population. Over time, lots of litter had built up in a certain patch (you could hardly see the ground).

I noticed that I felt a sense of annoyance every time I walked through the tiny valley of trash (not annoyance at humans, but at the interruption in what was otherwise a nice miniature nature walk). So one day I bought some surgical gloves and trash bags, put on a podcast, and cleaned the path. It took less than two hours to remove 98% of the litter (the rest being things like bottle caps that would have been laborious to track down and collect). 

The result: I got a clean path, a few hundred other people got a clean path, and I got to think of myself as "the kind of person who cleans up the commons," which was more personally satisfying than any donation I made that year (because I am irrational).

Although I agree with the message that fuzzy-feely  altruism can benefit your own well-being and motivation to do high impact altruism, I cringed a bit at the title. Please consider feeding only sterilised stray cats. Thanks ^^'

Otherwise, by feeding a fertile stray cat you contribute to creation of more starving, diseased cats. So even for the fuzzy feely altruism it is important to not just see what you want to see (the one moment of a pleased stray cat you interacted with), but assess relevant (future) consequences to not cause more suffering. 

Actually, sterilising would be the more valuable doing than feeding, but that again may not be a fuzzy feely altruism anymore, as the immediate reaction of the cat certainly will not be gratefulness. And we usually need (immediate) gratefulness of other beings to create the desired improved mental well-being for ourselves and motivation. 

There are many reasons that humans tend to do this, and I'm very familiar with them! I wrote part of my thesis on this topic.

Nevertheless, my feelings remain. The problem isn't ignorance, but (the concept I was trying to represent with "irrationality").

How much should conflicting desires to be locally kind and globally good affect our choices about living in EA bubbles, where our locally kind choices might multiply the effectiveness of effective people? I had previously felt it was a strong reason to live in an EA bubble, but perhaps this was due to stupid reasons.

I don't have much time to respond here and haven't thought much about my thesis since I wrote it almost seven years ago (and would probably find much of it embarrassing now in the light of the replication crisis + my better grasp on philosophy). A few notes:

• I think that humans do something akin to CEV as part of our daily lives — we experience impulses, then rein them back ("take a deep breath", "think of what X would say", "put yourself in their shoes"...) It seems like we're usually happier with our choices when we've taken more time to think them over (though this isn't always true — people do argue themselves out of doing what they later believe they should have done). I've always pictured CEV as an extremely advanced version of this, but one that we would still recognize as descending from that original process.
• I also see ethics as a subset of desires. 
• I don't think "System 1" and "System 2" are real things in a sense that would make "System 2 has no desires" a sensible statement. Humans do a lot of things with our minds, those things have properties, and we try to split them up based on those properties — but this doesn't mean that the resulting clusters have to refer to real things. Even when I'm doing my deepest "EA thinking", there's always  a thread of "why" I can pull that leads back to something I care about "just because", and it seems like the same thing would be true for every other mental process. (And by "a thread", I mean "multiple threads", and sometimes I'll look at where all the threads end and wind up needing to make an arbitrary about which "just because" thing to prioritize.)

I was using very casual language here, and there might be a better word than "irrational".

The complex concept I was casually representing: "It seems good to be someone who feels more satisfaction when they do more good for more people. This isn't how my own feelings of satisfaction work, which makes me less motivated to do more good for more people than I wish I were."

"Irrational" refers to the desire to feel a different way than I actually feel, with a hint of "this is especially awkward because I've had plenty of time to reflect on these feelings and try to change them". Maybe "unreasonable" is a better word, or even "imperfect".

I loved reading this. Thank you.

Thank you for writing this. I can relate well to the refreshing and restorative effect of small acts of kindness.

I think there are way too many narratives encouraging people to practice small acts of kindness that produce equally small benefits.

Thanks for helping me notice that I have one of those narratives floating around in my head without being questioned. Questioning it right now feels kind of sad, I really liked the idea that my small acts of considerateness will maybe potentially some day turn out to have been very important for the future of everything.