3 ms·
That may be true when it comes to product design (though I doubt it). Math is math. A car is a car... there's nothing racial or sexist about a car, unless it f
by TimTheTinker 5y ago
That may be true when it comes to product design (though I doubt it).
Math is math. A car is a car... there's nothing racial or sexist about a car, unless it fails to come with an adjustable seat.
- throwaway946513 5y agoWhich while you're not wrong - there are issues with some programs where it could unintentionally discriminate. The easiest visual instance of this can be photographic filters - Google's Pixel 6 emphasized better photos for people of Color, where other sensors/phones may not be correctly calibrated for PoC. This is a pretty minor instance but easily visualized. AI that does facial recognition or facial identification has in the past had issues with the faces of non-white people. It isn't a result of a 'racist' programmer. It's just from the fact that most testing and the data sets of these softwares is lacking the necessary diversity. It is part of including more people in software engineering is to combat these issues.
- TimTheTinker 5y agoAgreed - AI and training data definitely can have unfair bias.