You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Could the algorithms be adjusted to be less racially biased? YES! Could the data be adjusted to contain less racial bias? YES! It is a *dumb* argument because BOTH THINGS ARE SOLUTIONS. Can we talk about how tech can do better now?
If you are one of these "is it the data or the algorithm?" people, whether you are aware it or not, you are diverting energy away from an important discussion about real harms to real people to a pointless discussion of semantics.
This is a common behavior when people are confronted with the idea that a culture they care about and are involved in is racist. It moves the discussion from an uncomfortable conversation about racial bias to a more comfortable one about technical details.
People have been using this tactic to avoid discussions about anti-blackness for hundreds of years. The US founders punted on the question of whether black people were people, and thus deserving of the full rights and protections of constitution, by making the 3/5ths compromise.
Talk about turning a race problem into a math problem! So, if you're encountering a lot of strong pushback over this rhetorical manoeuvre about whether it's the algorithm or the data, it's because in 2020, nobody has time for you to catch up to the conversation.pic.twitter.com/k4HVs1UXCX
This is the one occasion where "it isn't a conscious being therefore it can't be racist" makes sense - but only in that you need to punt the racism back up to whoever created the flawed system as a whole. Whoever built it, built it wrong. How they fix it is a separate concern.
I don't know how 'data or algorithm' gains any traction. It's people. They collect data. They choose algorithms. They tend to ignore structural inequality. It's the diffusion of personal responsibility for work as an employee to an unaccountable and indifferent corporation...
Also I thought the whole value of machine learning is that the data becomes part of the algorithm.
This is me being white and coming from a different enough different society, but aren't these both conversations important and mutually-beneficial? The harms/effects discussion sheds light what needs to be changed for those of us that are not conscious enough about it
The semantics conversation helps identify flaws to work on. These flaws might be obvious to people that suffer from the biases, but need to be elaborated upon (presuming that they can't be made entirely explicit) for those whose intrinsic interests pull away from the conversation
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
