https://www.smithsonianmag.com/history/unraveling-the-colorf...

Thanks for that link, which confirms xdennis's post above is correct. A few lines from that article, emphasis mine:

> blue’s connection with boys and pink’s with girls wasn’t solidified until the mid-20th century... the gendered nature of the colors was either reversed or at best inconsistent (Paoletti’s original argument, which has since been misconstrued as a total reversal).

> ...5 million books published in the U.S. between 1880 and 1980, he found many examples of the standard “blue for boys” and “pink for girls” association, but virtually none for the reverse... “remarkable consistency in gender coding over time” rather than a sharp shift around the 1940s, leading him to conclude that the magazine excerpts often cited as proof of the reversal were anomalies.

> ...follow-up analysis... published between 1881 and 1930... recorded 34 instances of the standard color coding and 28 instances of the reverse... “Pink-blue gender coding showed a certain degree of inconsistency (though not a reversal)... “However, the true extent of that inconsistency is still unclear, as different kinds of sources return dramatically different pictures.”

@ternaryoperator: perhaps that's a good example of AI not being real-world-helpful, if you did ask it for "specific citations of pink being recommended for boys", because that sounds like it was asked to cherry-pick, and did.