I think this thought experiment can almost be summed up as “Are you the kind of person that ignores data and evidence which proves that your previously held beliefs are false?”
Whether it comes from an A.I. or from a scientist, some things are demonstrably true. Some people choose to ignore things that are true, and continue to live in a fantasy land. That will always be the case as long as humans are humans.
The benefit of an A.I. (at least one that is open source) would be that it won’t have any bias. It will digest all information it finds and draw conclusions. A task nearly impossible for any single person given the amount of information generated every single second.
This will lead inevitably to another road block: people will disagree with the morals and ethics of a machine. For example, an A.I. might come to the conclusion that all humans with certain genetic markers should be exterminated for the benefit of the human race. Perhaps the A.I. is correct according to all evidence, but will ANYONE ever agree with the conclusion?
In a world mired in a swarm of information and misinformation an A.I. will help cut through the noise, but there are some moral and ethical dilemmas that will never have a right or wrong answer.