Dan O
2 min readMay 1, 2020

--

Hurriya,

everyone is running a “auto matching algorithm” — you are too. You look for some distinctive features of every person and every object you encounter, and you use those features to index those items. this is how the human mind work.

In your case that algorithm will index off of your skin color, not because it is important, but simply because it is distinctive. If you were one of two people that always dressed in neon green I bet people would mix you up. Just because it is so salient.

So they are not attending any less to you, than they are to anyone else, but their feature matcher just naturally locks onto that feature. Likely it will lock on to this until there are many South Asian at your firm, then naturally that would go away.

Still you might decide that this is offensive, and all people should take extra care to override their natural matching algorithm so they avoid this kind of error. I offer no opinion on how important that is (I am a white guy, so its hard for me to relate to the feeling of being mislabeled), just know that this is a request to take special care in this case, relative to the care they take with others in managing their own matching algorithm.

but even if they do take this special care, and they never make this mistake again, sadly you should know that internally it is still the first indexing that they are doing, and likely will remain that way until that is no longer a salient feature. (There are studies of quick reactions to white and black faces… it is shocking just how baked in the associations are! and those associations exist for both blacks and whites too.) our mind is under much less direct control, than we imagine it to be….

Still, the question to ask, is really more about how you are treated by those that make this mistake. Are they more dismissive of you? kinda of hard to tell, but for me, it would be the more important question to ask.

— d

--

--

Dan O
Dan O

Written by Dan O

Startup Guy, PhD AI, Kentuckian living in San Fran

Responses (1)