And you think humans will do a good job at correctly assessing pieces of content on the sliding scale of “hate”? When one moment it’s fine debating gender issue

And you think humans will do a good job at correctly assessing pieces of content on the sliding scale of “hate”? When one moment it’s fine debating gender issues and a moment later it’s hate speech? For 35,000 people? This is an impossible problem, both for AI and humans. Any suggestion that this will be solved is pandering bullshit to the NYT crowd.