Yes, and I’d like to add a system that has racism baked in is much like a deep learning model with racism baked in. For the model you’d obviously want to compen

Yes, and I’d like to add a system that has racism baked in is much like a deep learning model with racism baked in. For the model you’d obviously want to compensate those prejudicial factors because they imply an incorrect bias with the model. It’s like if you trained your hot dog detector with only pictures of hot dogs with buns. It’s not going to recognize hot dogs for what they truly are so you’d have to correct the input bias to fix it.