> But when you ask it to follow a chain of thought and generate it in the output, the eventual conclusions it lands on is scarily human I have tried this so ma

> But when you ask it to follow a chain of thought and generate it in the output, the eventual conclusions it lands on is scarily human I have tried this so many times for novel problems and no matter what I do, it eventually recognizes what we're trying to solve and determines it's unsolvable. Since nothing new comes out I can only attribute the times it does seem to do symbolic manipulation as regurgitation.