This is just scary, there's no way this will be unbiased.

Sure, some stuff people write is just hate with the only motive to offend others.

But there is also facts that can be viewed as offensive, while still being true. Does that mean facts will be hate speech and automatically hidden by an AI? Ouch.