I'm excited to see future improvements to AGI. Everytime I debate GPT on a subject I ask it to defend with its life, I still win without much effort, and that isn't because my position is flawless. I suspect in a year or 2, this will no longer be the case.