there's a selection bias at play here

suppose you need to either be in the top percentile of handsome and the top percentile of emotionally aware to get a job as a hollywood actor, but these attributes are uncorrelated in the general population, so 99% of top-percentile-of-handsome people will be top-percentile-of-sensitive and vice versa

then the pool of hollywood actors you observe will contain, out of every 199 people, 99 who are top-percentile-of-handsome but not of sensitivity, 99 who are top-percentile-of-sensitive but not of handsomeness, and 1 who is both

someone observing this result but not understanding the process that led up to it might think that emotional sensitivity makes you ugly or that being handsome makes you emotionally oblivious, even though (by hypothesis) the traits are uncorrelated. in fact, this negative correlation in the selected group can survive even a fair bit of positive correlation between the traits in the general population

similarly, it's easier to get a programming job if you have a history of delivering successful products, or if you have a lot of line-by-line coding acumen