Ohai, Am 15. May 2019 um 12:59:08 +0200, schrieb Jay Vaughan:
"it works for erveryone, no matter what" = that naive that I will not even comment on this
Enlighten me - how is believing that technology should work for everyone, no matter what, considered being naive? Since I believe this hasn't been answered: Technology is created by people, and therefore shaped by their bias. This doesn't have to be done out of malevolence; implicit assumptions are enough.
If a soap dispenser is only tested by white people, chances are that it won't register hands of (some) PoC. If medical studies only include amab people, then the equipment designed using data from those studies might do unforeseen weird shit when used on afab people, and unforeseen weird shit and medical equipment are a bad combination. If a chatbot is trained on sample texts written by a broad sample of people living in a sexist society, guess what happens? Now I kinda like your idea of technology as an equalizer, nonetheless - it is true that the building blocks themselves are usually simple enough to be unbiased. But then the question is how they are put together, and if we want to use technology as an equalizer, we need to be mindful of how people enter the equation. I believe that technology has both the potential to safeguard against bias and the potential to amplify it. When there is sufficient input from a diverse group of contributors that eliminates implicit assumptions and design decisions that exclude people, it can be the impartial and impersonal force you imagine. But if this is not the case, it only makes the bias worse in the way I described earlier. If you want tech to be unbiased, you should be in favor of increasing representation for that reason alone. cheers, Thomas