This is about confidence not truth, confidence is built into ML, it's how it works... the problem is these solutions currently just quantize that output into a

This is about confidence not truth, confidence is built into ML, it's how it works... the problem is these solutions currently just quantize that output into a single result and obscure how close to the threshold of another possible answer it was.