I've seen a lot of articles making this suggestion, but never do they seem to actually check if the software was indeed written by white men. It's as if they don't want their biases contradicted by evidence.
Who the biased algorithms are written by are pretty irrelevant. The argument is they are biased against anyone who is not a white male, and that we should be aware of that.
The article has a hypothesis right in the headline: "algorithmic bias is a result of the insular, mostly white tech industry". So no, if it was written by an entirely Indian staff at an Indian outsourcing company, that would discredit the idea.
I agree in the sense that it’s lazy writing. But have taken it to be cultural, not racial, in the sense that technological biases against white men (regardless of the identities of builders of said technology) do not threaten their identity in the same way that journalism might be threatening to powerful men in an age of massive social change. But I’m not a white man so that’s just my projection. It would be good to see a list of AI projects or UI patterns or tech startup trends articulated by white men that feel threatening to their identity to round out the larger conversation around bias.
Because in our society people of different races are treated differently.
If you don't realize this, you either need to spend more time interacting with people outside of your bubble, or your friends aren't comfortable enough with you to share their experiences.
If neither of those are the case, you should tell us where you live, and what the secret sauce to this magical, post-racism colorblind Utopia is.
The funny thing is that the valley is doing everything in its power to get rid of these "biases", even when they're proven out by machine learning, empirical evidence, and profitability.
For instance, Facebook says that it will automatically stop assigning targeted ads by race, because it could get the race wrong. [1]
However, when they got the race "wrong" even though the person might have interests stereotypically associated with those races, there was a big outroar. If they had instead called the race groupings by some other name such as "urban" for rap and basketball, for instance, I think that would not be so much of a big deal.
I encourage people to look at this problem as something that needs to be solve and not an assault at their ignorance or character. Instead of assigning or deflecting blame, perhaps it's best to just be aware of the implicit biases in technology and using "historical data" to predict the future. That is, relying on historical data and assigning values to predictions as a result may be harmful to the growth of those negatively affected by such values. Especially when we have far from perfect knowledge and so many of the "sources" we rely on are also inherently biased.
Imaging looking at our history through the colonized vs. the colonizers. Do you think you'll get the same account?
11 comments
[ 2.6 ms ] story [ 37.0 ms ] threadIf you don't realize this, you either need to spend more time interacting with people outside of your bubble, or your friends aren't comfortable enough with you to share their experiences.
If neither of those are the case, you should tell us where you live, and what the secret sauce to this magical, post-racism colorblind Utopia is.
For instance, Facebook says that it will automatically stop assigning targeted ads by race, because it could get the race wrong. [1]
However, when they got the race "wrong" even though the person might have interests stereotypically associated with those races, there was a big outroar. If they had instead called the race groupings by some other name such as "urban" for rap and basketball, for instance, I think that would not be so much of a big deal.
[1] https://www.nytimes.com/2016/11/12/business/media/facebook-w...
Imaging looking at our history through the colonized vs. the colonizers. Do you think you'll get the same account?