Personally, what I would want, is for the model to predict the most likely outcome of any political choice or policy, based on the vast training set and learnin

Personally, what I would want, is for the model to predict the most likely outcome of any political choice or policy, based on the vast training set and learning it encoded. Where I think the AI should remain neutral is when deciding what outcomes are desired. That's inherently human. Say you want to pay no taxes and don't care about people poorer than you. You wouldn't want it to patronize you or try to convince you otherwise here. But for any given political platform, policy, and so on, you'd want to know what it predicts would be the outcome related to your goal, would it most likely result in me paying less tax or not, at what cost, what else would I lose, etc. I wouldn't want it here to be neutral about all proposed ideas for say improving the economy and lowering taxes. I need it to be an AI and actually provide predictions.