brace yourself (to prevent further injury)

That's what I meant with the last bit about scoping, yeah.

I think worrying about becoming a "reply guy" is reasonable, but I'll let you in on a secret that will hopefully put you at ease: the core philosophy of the reply guy ethos is "I'm gonna make this about me!" When you're genuinely appreciating or engaging with something in earnest, you will never do that accidentally. Also, I don't speak for everyone but for me, regular and reliable commenters are my favorite thing in the whole world.

This is an extremely weird take. You're really leaping to conclusions about the consequences of those choices and frankly I don't like your energy. Goodbye!

i think the feeling about the necessity of metrics to weed out hate comes from a valid fear especially with how twitter moderates (barely), but here there has been a pretty explicit promise (and practice!) of proper humans-actually-reading-the-reports moderation

the punishment for posting hate is getting posts deleted and being banned, not being ratioed and the post staying up, and i think unless or until staff proves incapable of effective moderation what we have now is the better option

I don't follow how not being able to give unspoken approval to Good Posts could lead to the promotion of hateful posts. That seems more like a site culture thing; people would have to be actively choosing to circulate hateful content (and failing to report it).

In my experience a system with visible likes doesn't do much to discourage people being assholes unless there's also a negative reinforcement mechanic like Reddit downvotes, but even then what it measures is usually group social approval, and that is highly localized and can easily backfire. For example, Elon Musk's tweets usually have more likes than they do comments (which tend to be negative, at least at the top) or retweets (could be positive, negative, or neutral), or quote retweets (commonly the form a negative retweet takes). What can we conclude about a guy who everyone seems to be universally calling the world's biggest dipshit when they type text and post it, but who also has upwards of 100k likes on some of his most dipshit opinions? On a smaller scale, what can we conclude about a terf post that hasn't circulated beyond a terf subreddit and thus has hundreds of approving upvotes and comments and only a few dissenting comments, all of which have been downvoted into invisibility?

I personally think a lot of the problems with social media sites have been the direct result of building systems to order and regulate feedback; it typically means that human beings still have to manage the feedback but must do it by interacting with the system instead of with each other directly. The concept of a "ratio" is exactly what this is; if you know that an opinion's level of social acceptability is going to be judged at a glance, then a post having far more comments or quote retweets than likes is the visible-at-a-glance indicator of, "oh, a lot of people think this opinion sucks." And it's half-ironic; the only reason to play this game is to try to hit someone with enough negative feedback that they delete it, because even commenting to say it's a shitty opinion expands its reach.

I don't know if no visible likes or no likes on comments is a solution to this--I'm generally wary of any "people will do x if we prevent them from doing y" system in tech, although it doesn't seem like the Cohost staff are operating from that perspective in the way that causes major problems. But it is an opportunity to try something else for a while and see if new tools lead to better fixes.

I think the important thing here is that cohost doesn't expose the number of shares or likes something has, so you can't really number watch an individual chost without taking great pains, and I love that! But if you could generate an "ack" notification on a comment for the receiver like you can heart a post—a silent nod between you and the commenter—I think there's good merit to that