Right now everything seems to travel to the intersection of "Free Speech" and "Dangerous speech", creating a massive pile up of conversations going no where. A

Right now everything seems to travel to the intersection of "Free Speech" and "Dangerous speech", creating a massive pile up of conversations going no where. Arguably this logjam resolves itself if you assume: 1) Everything is content, True/false, wrong/right being labels we use after consuming it. 2) All our social media, every new type of it, all UI/UX, payment models, subscriptions, memberships - simply Amplify or Suppress the propagation of content. This makes our tech discussions relatively value agnostic, and moves the question to whether i) the use of these tools was just ii) The extent of force used was just So yes, all our major firms have this power. ANY firm can have silencing or amplifying power. The question is how we define the norms, and the empirical evidence we generate that shows us what choices are better than others. A step below that is the empirics and informational tools we have to do the science. I used my masters dissertation as an excuse to explore whether a particular rule change on a forum worked. Honestly while it was fun to reach some conclusion, there are major gaps in our toolkits to explore these issues. Sentiment analysis corpuses for non-english corpuses are definitely one of the data sets we need more off, and we need ways to constantly update these dictionaries. Dealing with code mixed language is another problem. I'm sure someone has solved this, one of the issues is precisely that cross talk on experiments is low and every discipline is looking at this space. But that is something you can invest your time in if creating this evidence matters to you.