I don’t think the author’s OVERcompensating. There may come a time when little girls are insufficiently told that they look adorable, but I seriously doubt that today is that time, nor that we have happened to stumble through memetic evolution onto the exactly correct amount to be complimenting little girls on their looks.
My own take is that if this is a little girl you see a LOT of (like, say, your daughter) then you might occasionally congratulate them on looking nice (especially if you know that they DID put some effort into it), but as a random stranger operating on timeless decision theory, the author’s advice is pretty sound.
(Edit: and no, you’re not being dishonest by telling them the first thing you notice, any more than you’d be lying to refrain from telling a below average child that they are stupid or ugly when you first meet them)
but as a random stranger operating on timeless decision theory, the author’s advice is pretty sound.
Your opinion makes sense to me, but I don’t really understand how it involves TDT. My thoughts would probably be more along the lines of “not enough people talk to little girls this way, and it would probably help them be happier in the future if I did so.”
A person puts a lot of effort into something, you notice the good job they have done, but refuse to let them know that you have noticed it, and justify your actions with the (not-yet-formalized) TDT?
I believe that my post on blind spots would be relevant here, sorry.
My prior on cute little girls is that their parents put them in a dress, and that their work mostly consisted off putting up with it. (“Little girl” is ambiguous—as they grow up the prior on how much effort they’re personally putting in obviously goes up).
I believe there are negative consequences to to overcomplimenting them on looks, same way there are negative consequences to being told you’re smart. There may very well turn out to be consequences of being told too often that you’re hard working.
There are also, I’m sure, negative consequences to not being told you’re pretty often enough. A perfect agent probably spends some time figuring out how to tell when young girls looked good due to their own efforts, compares how often they instinctively compliment that over other ways they might begin the conversation (perhaps taking note of how often they begin conversations with young boys who look particularly well dressed with things like “Who’s this handsome little man!”) and then randomly decide whether to do so. Even without regarding how the rest of society also will impact the child, I doubt the ideal number comes out to more than 1 in 6 or so for the average person.
If you don’t have time to do all that, I think beginning conversations with “how are you?” and going from there is a perfectly good heuristic. If the girl is particularly proud of her clothes, I suspect it’ll come up, and you can compliment her then.
True, but it can still be a trial for someone that young.
I doubt the ideal number comes out to more than 1 in 6 or so for the average person.
I strongly suspect that you made this number up on the spot. If not, I would be interesting to hear how you arrived to it, and what the error bars are.
Oh the 1 in 6 number was totally made up, and was anchored by thinking about randomization which made me think about dice which made me think about 10 sided dice (because I’m a gamer and that’s what comes first time mind), which made me think about “1 in 100” chances as compared to “1 in 10″ chances, and the latter seemed like a reasonable upper bound on how often you shoud compliment a girl on her looks (based on intuition), and then in case 1 in 10 was still too high I went up to the next most recognizeable dice-size, which was 1-6.
I’m NOT a perfect agent with infinite time, and I haven’t thought about this particular instance of this problem before today. But I suspect the upper bound for how often girls need complimenting on their looks from strangers is somewhere in that order of magnitude. I don’t think that there’s much useful ground between using your intuition and doing a bunch of intensive research.
I could see the answer ranging from 1 in 20 to 1 in 6ish (I’m trying to imagine 1⁄5 as a viable answer and failing, but at this point I’m hopelessly anchored). That’s in the world where everyone in society is trying to do this at once, as opposed to a minority of agents operating against huge cultural biases.
This does not sound believable to me. I very much doubt that a perfect agent would find these sorts of questions to be well-defined. It sounds like you’re talking about the likelihood of a bulkhead failure or something completely technical … when this is not technical at all.
The amount of input my son puts into choosing his clothes is zero (because he’s not even two). When he’s older, the proportion of the clothing selection process that’s based on his input is highly likely to increase. An agent can’t measure or estimate the proportion for a particular child? An agent can’t (or shouldn’t) change her interaction with a child based on the estimate proportion?
I don’t think the author’s OVERcompensating. There may come a time when little girls are insufficiently told that they look adorable, but I seriously doubt that today is that time, nor that we have happened to stumble through memetic evolution onto the exactly correct amount to be complimenting little girls on their looks.
My own take is that if this is a little girl you see a LOT of (like, say, your daughter) then you might occasionally congratulate them on looking nice (especially if you know that they DID put some effort into it), but as a random stranger operating on timeless decision theory, the author’s advice is pretty sound.
(Edit: and no, you’re not being dishonest by telling them the first thing you notice, any more than you’d be lying to refrain from telling a below average child that they are stupid or ugly when you first meet them)
Your opinion makes sense to me, but I don’t really understand how it involves TDT. My thoughts would probably be more along the lines of “not enough people talk to little girls this way, and it would probably help them be happier in the future if I did so.”
The initial statement was off the cuff and not really well thought out. The followup statement in the next comment down is probably more accurate.
A person puts a lot of effort into something, you notice the good job they have done, but refuse to let them know that you have noticed it, and justify your actions with the (not-yet-formalized) TDT?
I believe that my post on blind spots would be relevant here, sorry.
My prior on cute little girls is that their parents put them in a dress, and that their work mostly consisted off putting up with it. (“Little girl” is ambiguous—as they grow up the prior on how much effort they’re personally putting in obviously goes up).
I believe there are negative consequences to to overcomplimenting them on looks, same way there are negative consequences to being told you’re smart. There may very well turn out to be consequences of being told too often that you’re hard working.
There are also, I’m sure, negative consequences to not being told you’re pretty often enough. A perfect agent probably spends some time figuring out how to tell when young girls looked good due to their own efforts, compares how often they instinctively compliment that over other ways they might begin the conversation (perhaps taking note of how often they begin conversations with young boys who look particularly well dressed with things like “Who’s this handsome little man!”) and then randomly decide whether to do so. Even without regarding how the rest of society also will impact the child, I doubt the ideal number comes out to more than 1 in 6 or so for the average person.
If you don’t have time to do all that, I think beginning conversations with “how are you?” and going from there is a perfectly good heuristic. If the girl is particularly proud of her clothes, I suspect it’ll come up, and you can compliment her then.
True, but it can still be a trial for someone that young.
I strongly suspect that you made this number up on the spot. If not, I would be interesting to hear how you arrived to it, and what the error bars are.
Oh the 1 in 6 number was totally made up, and was anchored by thinking about randomization which made me think about dice which made me think about 10 sided dice (because I’m a gamer and that’s what comes first time mind), which made me think about “1 in 100” chances as compared to “1 in 10″ chances, and the latter seemed like a reasonable upper bound on how often you shoud compliment a girl on her looks (based on intuition), and then in case 1 in 10 was still too high I went up to the next most recognizeable dice-size, which was 1-6.
I’m NOT a perfect agent with infinite time, and I haven’t thought about this particular instance of this problem before today. But I suspect the upper bound for how often girls need complimenting on their looks from strangers is somewhere in that order of magnitude. I don’t think that there’s much useful ground between using your intuition and doing a bunch of intensive research.
I could see the answer ranging from 1 in 20 to 1 in 6ish (I’m trying to imagine 1⁄5 as a viable answer and failing, but at this point I’m hopelessly anchored). That’s in the world where everyone in society is trying to do this at once, as opposed to a minority of agents operating against huge cultural biases.
This does not sound believable to me. I very much doubt that a perfect agent would find these sorts of questions to be well-defined. It sounds like you’re talking about the likelihood of a bulkhead failure or something completely technical … when this is not technical at all.
The amount of input my son puts into choosing his clothes is zero (because he’s not even two). When he’s older, the proportion of the clothing selection process that’s based on his input is highly likely to increase. An agent can’t measure or estimate the proportion for a particular child? An agent can’t (or shouldn’t) change her interaction with a child based on the estimate proportion?