@timtimtim

I think the black box issue I am trying to discuss here is more on questioning the process. How they train the model, if the model contains more bias, then it should do. For example, scaleAI may have their own data labeling standard, which could intentionally create more bias.

1 reply

0 recast

0 reaction