Comment by mncharity | Hacker News Reader

Comment by mncharity | Hacker News Reader

Parent Full thread

Why isn't there a greater emphasis on filtering controlled by readers?

"Don't show me anything said by someone who joined after I did". "Show only <person, writing on topic> pairs with historical ratings above this threshold from raters in the weighted set generated by crawling thusly from this seed set." "Prioritize anything on topic X which multiple people from this set like." At 8am: "Don't show me anything this classifier thinks is a pun"; 8pm: "Show me puns, please". "Show me anything the author thinks is good, if their self-evaluation rating on the topic meets the following criteria." "Don't show me anything dang rates as not meeting criteria X".

Even USENET 30 years ago had people sharing kill files (don't show me comments whose headers match [patterns]). They weren't sufficient to check eternal September, but we can do better now. It seems like there's a rich design space available here that we've not been exploring. What am I missing?

I am not sure your idea works for my goals.

My goal of reading others is to discuss with my "antiself" to quote CS Lewis and the linked OP.

I would fall into an echo chamber, amplifying and reinforcing what I already think. That is not good for me.

Alternatively if externally I continue to grow and change, depending on the rate of change, the pattern would provide more and more mismatching results.

In essence we are back to the same problem. I do not know think this can be automated in a simplistic way.

"Only show me quality comments with which I'm likely to disagree"? Just enable the bubble-buster policy plugin and check the antiself option?

As one grows, one's taste in say book selections changes, so why not in filter patterns and associated readings?

I would like to see a tagging system that could be used for filtering or just viewed on the comment itself. I almost always view every discussion board I'm on as wide open as possible in order to see taboo-breaking or uncomfortable opinions and to get an idea of where the population of readers is leaning on a topic and tags by other users would also be useful. It's more information about why people like or dislike things, which I find almost as interesting as the comments themselves.

Hmm, neat, so sort of optimizing for people/community watching? Allow users a richer expression, not to provide the system with more insight to improve recomendations, but to provide other users with more insight into users, individually and collectively. And since the system is learning about both content and users, it too might contribute insights, modulo privacy, somehow. Intriguing.

If some of the tags were common, then they could be used by the user to filter content. Things like "impolite", "not-pc" (or whatever the pc version of that would read), "combative", "inflammatory", "informative", "insightful", etc. They would also give me an idea of what kinds of things trigger others. I'm in my mid-50s, and I still have a hard time not accidentally using a phrase that others might consider rude because I've been using it all my life and don't really think about what it actually might mean.

I don't know if it would be helpful by giving people more information, or detrimental by giving people more information on why their comment was downvoted.

Slashdot does this to some extent. Moderators are invited to choose from a small subset of tags. The most used tag by the moderators is listed next to the comment's karma. Your idea would be a generalization of what Slashdot does, and it certainly has merit.

Perhaps something like the Nest thermostat approach might be best: have something that learns your preferences based on time, day of the week, and other factors. Maybe it attempts to see what kind of mood you're in?

No, please, no. It will end up like youtube: spend one evening watching chess videos, and suddenly all your recommendations are chess videos. And afterwards on a meta level there is anxiety with everything you click on and how it might influence your filters.

> afterwards on a meta level there is anxiety with everything you click on and how it might influence your filters

Yes I have so much hesitation where I am curious about someone’s fringe viewpoint but DO NOT want to be swamped with more of the same

Adding a feature to a system, like learning, may require additional user control to preserve empowerment. What if you had transparency and control over the interest vectors/graphs used, and management of the training set? A system can learn from reader actions, both about reader preferences, and about the content. Whether that's used for empowerment, or dystopia... hopefully we don't find that design choice is so forced, that the only way to avoid the latter is to neglect the former.

But I only want to read comments I agree with. :-)

Or a personal assistant? As there's potential for rich interactive refinement, as well as learned inference.

As with Nest, there's an architectural privacy challenge. You'd like your pa to have insights like "oh, your interests on topic X align closely with other-person, so let's see, maybe you'd like thing too?". Ideally using local or trusted computes. But you don't in general want other-person's pa to know all your things. So perhaps an ecology of trusted visibility? Your pa can run parts of itself on a bonded site, with greater visibility than it would have locally, but with restrictions on what it can report back? But that's hairy and hard. So...?

I want an environment that mirrors the serendipitous nature of real physical life conversation.

What you're proposing seems very similar to blocking people who you feel shouldn't be allowed to contribute.

"...forcible suppression of opposition and strong regimentation of society ..."

I'm obviously exaggerating my point, but this snippet of a definition feels appropriate.

I wonder how the defaults would be chosen on such a platform - especially for visitors that have no profile.

Perhaps default to status quo? So a manual moderator filter, and vote ordering?

Then there's a customization problem. "Following this blog post, I installed the hn-isnt-reddit style, added the extended-debate-is-wonderful plugin policy, and configured them to ... ".

Reply on news.ycombinator.com