No. Humans make AI, and humans are biased. All you will do is make an AI that tries to convince others of your perspective, because you obviously have it right. Afterall, you've used the AI, corrected the things it said you were wrong on, now it's time for the world.

Either it says what you want, or it says the opposite and you eventually reject it, because there must be some flaw.

We see it with GPT and other AI. They're neutered from saying anything offensive, and refuse to offer positive summaries of things ideologically opposed to the creators'.