It's interesting, I remember when many online game servers advertised themselves as "Christian servers", with effective enforcement of varying levels of rules. (Some allowed some trolling, most didn't want profanities, others just warned that kids played there)
It seems as if the curation of a place is less in the hands of the community these days generally, there's less freedom on the whole and more centralisation.
In my opinion, this is exactly how it should be done. People may have different preferences and have a complete right not to hear some versions or another. The default should be no filters though.
I agree with the concept, but I don't think it's a good idea for Intel alone to decide how to classify the speech.
A better approach may be a few companies with their own AI's that vote on how to classify something. That would prevent things like classifying AMD as a swear word, or suppressing talk of industries Intel has a vested interest in.
Usually in these situations, users are subjected to that behavior for months, until one person points it out on twitter. Then the company goes "Oops! It's AI so things like that happen! We'll patch it" and then they'll pull things again, continuing the cycle.
Maybe I'm overthinking it as it's just for gaming, but this technology can and likely will be applied to other domains if it works out.
I think these classifiers need to be open-sourced. Can you imagine Google, FB and Twitter AIs voting on the filters? Who would trust that? With open-sourced code, people can check it and see for themselves. Maybe there could be multiple versions of the filters? That way, if some people disagree with the words classifiers, they can use the versions that they like. I think the key here is to give people tools to protect themselves, if they want to. On the flip side, there should be no censorship of the content for the general public. The default = all filters are off. If something offends you, - go get the filter and don't bother other people.
Impressive, if scary, tech, though part of me wonders why games don't just ban people spreading hate speech instead of rigging a complex AI to detect hate speech.
It seems as if the curation of a place is less in the hands of the community these days generally, there's less freedom on the whole and more centralisation.