When you call for violence against non-combatants you're breaking the law in every single western country. If there were only one web browser and the company behind it were implementing universal blocking measures maybe I'd agree with you, but honestly I'd have to think long and hard first. Radicalisation is impossible to survive in the long run as the average power an average individual keeps going up.
I find it an interesting question because:
A) Not every video branded as culturally unacceptable will be. Not every video is as bad as the worst-case hypothetical used to justify the content classification.
The landscape of cultural attitudes differ from California-based content minders. The categorization can be flat out wrong, there will undoubtedly be a small percentage of videos that even the minders see as mis-classified.
B) Social interventionist policies can - and often do - backfire.
e.g.: Teens that deliberately seek out taboo. The allure of R movies, M games, Explicit Lyrics, and underage binge drinking can cause them to live a period of their life less well-adjusted than if that content wasn't aggressively filtered from their lives in the first place.
>videos that contain inflammatory religious or supremacist content
It's very easy to claim content is inflammatory or supremacist . This will be highly subjective, which is the problem.
I personally know people here in the bay area that would have no problem labelling lots of campaign talk by Trump with those tags.
also they have to determine these norms for the whole planet (without North Korea), now every attempt so far that tried to set cultural norms for the whole world has failed, lets see if they do better.
That's not entirely true. Abstract advocacy of illegal violence is protected speech under Brandenburg v. Ohio, 395 U.S. 444 (1969). Only when the incited violence is imminent (as opposed to at some indefinite future time) does the speech fall outside the bounds of the First Amendment.