Regarding AI bias and change. Some people can change more easily than others. The advantage ML models have is in being able to load test on a variety of scenarios in a way that's just not possible with humans. Changing bias is important, but being able to significantly test for bias before classification is often overlooked in the value it provides to reducing bias in society.
This is a great point and thanks for commenting. During the show, when I made the comment about humans being able to change more easily than data sets, I was thinking about your point that some people can change more easily than others. But you, and Dr. Sejnowski, are correct!
Regarding AI bias and change. Some people can change more easily than others. The advantage ML models have is in being able to load test on a variety of scenarios in a way that's just not possible with humans. Changing bias is important, but being able to significantly test for bias before classification is often overlooked in the value it provides to reducing bias in society.
This is a great point and thanks for commenting. During the show, when I made the comment about humans being able to change more easily than data sets, I was thinking about your point that some people can change more easily than others. But you, and Dr. Sejnowski, are correct!