News for hackers. Stuff that matters.
Re:AI models still racist, even with more balanced training by emteycz on Sunday May 01, 2022 @08:00AM (#31224308)
Or, the community knows that models will always have biases, and so the community makes its software work around that limitation by other means than modifying the model, because that might not be possible at all.
Re:AI models still racist, even with more balanced training by krapp on Sunday May 01, 2022 @08:34AM (#31224499)
The community that actually understands this as a problem and attempts to resolveit is not the community to which I was referring.