# AI and Truth Detection By:: [[Brian Heath]] 2023-01-19 The other day I heard someone say that in the future AI will be able to determine what is true and what is fake. From what I could gather, the person was concerned that AI could endlessly generate real and fake information to the point that only AI would be able to figure out what was true. This discussion went on for some time, but it speaks volumes about how little people understand truth and reality. Most of us seem convinced that truth is objective and singular and that there is a right answer. Nothing could be further from the truth (irony intended). Let's look at this a few different ways. First, do you know anyone who believes something that you don't? If so, how do you know that you are right and they are wrong? Can you read their mind, live their life, and check their brain chemistry and structure? Delusions are powerful, and we only think others are "crazy" because there aren't many of them. However, there are plenty of times when the majority was the "crazy" ones. See Germany in World War II. My truth is not your truth, even when we agree. Second, for those who believe in infallible science and mathematics, Godel's Incompleteness Theorem fundamentally shows you can't prove anything true outside of the axiomatic assumptions you make. Thus, you can't prove your assumptions are true, but only the mathematics within those assumptions are true. It's the same problem. My assumptions and life do not correspond to yours. Truth is elusive. Whoever tries to program an AI to determine truth from fiction without understanding the philosophical, scientific, and mathematical limits of knowledge is in for a world of hurt. Anyone who says they have figured it out is full of it. Don't buy the hype. #### Related Items [[Artificial Intelligence]] [[Truth]] [[Knowledge]] [[Philosophy]] [[Science]]