The alternative is building Functional Intelligence process flows from the ground up on a foundation of established truth? If 50% of training data is not factu

The alternative is building Functional Intelligence process flows from the ground up on a foundation of established truth? If 50% of training data is not factually accurate, this needs to be weeded out. Some industries require a first principles approach, and there are optimal process flows that lead to accurate and predictable results. These need research and implementation by man and machine.