I just posted a big effortpost and it may have been consigned to total obscurity because I posted it at the wrong time of day. Unsure whether I actually want the recommendation algorithm to have flattened time-discounting over periods with less activity on the site, or if I should just post more strategically in the future.
I have found the dialogues to be generally low-quality to read. The good ones tend to be more interview-like—“I have something I want to talk about but writing a post is harder than talking to a curious interlocutor about it.” I think this maybe suggests that I want to see dialogues rebranded to not say “dialogue.”
(Note, I don’t think it’s because it was posted at the wrong time of day. I think it’s because the opening doesn’t make a clear case for why people should read it.
In my experience posts like this still get a decent amount of attention if they are good, but it takes a lot longer, since it spreads more by word-of-mouth. The initial attention burst of LW is pretty heavily determined by how much the opening paragraphs and title draw people in. I feel kind of sad about that, but also don’t have a great alternative to the current HN-style algorithm that still does the other things we need karma/frontpage-sorting algorithm to do)
The initial attention burst of LW is pretty heavily determined by how much the opening paragraphs and title draw people in. I feel kind of sad about that, but also don’t have a great alternative to the current HN-style algorithm that still does the other things we need karma/frontpage-sorting algorithm to do)
It’s hard to envision a different solution to this problem. When I browse a feed and decide what to read, of course things like author, karma, title, and first paragraph are the things that determine whether I’ll consider reading. How else could things work?
@Charlie Steiner: Also see this comment thread on why it’s so important to pay outsized importance to stuff like the title and presentation. Excerpts from my comment:
if you can’t think of a way to present a piece of content such that people want to click on it, then they won’t click on it, and then all the work that went into making the high-quality content went down the drain.
And:
a bunch of things have to line up for an article to go viral: someone has to click on your content (A), then like it (B), and then finally follow a call to action like sharing it or donating (C). From this perspective, it’s important to put a significant fraction of one’s efforts on quality (B) into efforts on presentation / clickability (A).
(Side note: If this sounds like advocacy for clickbait, I think it isn’t. The de facto problem with a clickbaity title like “9 Easy Tips to Win At Life” is not the title per se, but that the corresponding content never delivers.)
I think time of day combined with when it was approved for front page can easily make all the difference between takeoff and just fading into obscurity.
This is an unfortunate situation, but I don’t have a solution.
I do wonder why posts with AI tags aren’t on front page automatically without human review.
Ah, that makes sense. I never see AI stuff on the front page or in recent discussions that isn’t worth at least a glance, but that’s a good thing. I do not want to see every little AI news piece on the front page.
What about leaning into the word-of-mouth sharing instead, and support that with features? For example, being able to as effortlessly as possible recommend posts to people you know from within LW?
Not crazy. I also think doing things that are a bit more social where you have ways to recommend (or disrecommend) a post with less anonymity attached, allowing us to propagate that information further, is not crazy, though I am worried about that incentivizing more groupthinking and weird social dynamics.
I’m not sure what the current algorithm is other than a general sense of “posts get promoted more if they’re more recent,” but it seems like it could be a good idea to just round it all up so that everything posted between 0 and N hours ago is treated as equally recent, so that time of day effects aren’t as strong.
Not sure about the exact value of N… 6? 12? It probably depends on what the current function is, and what the current cycle of viewership by time of day looks like. Does LW keep stats on that?
I have found the dialogues to be generally low-quality to read.
I think overall I’ve found dialogues pretty good, I’ve found them useful for understanding people’s specific positions and getting people’s takes on areas I don’t know that well. My favorite one so far is AI Timelines, which I found useful for understanding the various pictures of how AI development will go in the near term. I liked How useful is mechanistic interpretability? and Speaking to Congressional staffers about AI risk for understanding people’s takes on these areas.
I just posted a big effortpost and it may have been consigned to total obscurity because I posted it at the wrong time of day. Unsure whether I actually want the recommendation algorithm to have flattened time-discounting over periods with less activity on the site, or if I should just post more strategically in the future.
I have found the dialogues to be generally low-quality to read. The good ones tend to be more interview-like—“I have something I want to talk about but writing a post is harder than talking to a curious interlocutor about it.” I think this maybe suggests that I want to see dialogues rebranded to not say “dialogue.”
(Note, I don’t think it’s because it was posted at the wrong time of day. I think it’s because the opening doesn’t make a clear case for why people should read it.
In my experience posts like this still get a decent amount of attention if they are good, but it takes a lot longer, since it spreads more by word-of-mouth. The initial attention burst of LW is pretty heavily determined by how much the opening paragraphs and title draw people in. I feel kind of sad about that, but also don’t have a great alternative to the current HN-style algorithm that still does the other things we need karma/frontpage-sorting algorithm to do)
It’s hard to envision a different solution to this problem. When I browse a feed and decide what to read, of course things like author, karma, title, and first paragraph are the things that determine whether I’ll consider reading. How else could things work?
@Charlie Steiner: Also see this comment thread on why it’s so important to pay outsized importance to stuff like the title and presentation. Excerpts from my comment:
And:
Yeah, fair enough.
I think time of day combined with when it was approved for front page can easily make all the difference between takeoff and just fading into obscurity.
This is an unfortunate situation, but I don’t have a solution.
I do wonder why posts with AI tags aren’t on front page automatically without human review.
I mean, many posts with AI tags don’t meet frontpage norms. For example AI news isn’t timeless, and as such doesn’t make it onto the frontpage.
Ah, that makes sense. I never see AI stuff on the front page or in recent discussions that isn’t worth at least a glance, but that’s a good thing. I do not want to see every little AI news piece on the front page.
What about leaning into the word-of-mouth sharing instead, and support that with features? For example, being able to as effortlessly as possible recommend posts to people you know from within LW?
Not crazy. I also think doing things that are a bit more social where you have ways to recommend (or disrecommend) a post with less anonymity attached, allowing us to propagate that information further, is not crazy, though I am worried about that incentivizing more groupthinking and weird social dynamics.
I’m not sure what the current algorithm is other than a general sense of “posts get promoted more if they’re more recent,” but it seems like it could be a good idea to just round it all up so that everything posted between 0 and N hours ago is treated as equally recent, so that time of day effects aren’t as strong.
Not sure about the exact value of N… 6? 12? It probably depends on what the current function is, and what the current cycle of viewership by time of day looks like. Does LW keep stats on that?
I think overall I’ve found dialogues pretty good, I’ve found them useful for understanding people’s specific positions and getting people’s takes on areas I don’t know that well.
My favorite one so far is AI Timelines, which I found useful for understanding the various pictures of how AI development will go in the near term. I liked How useful is mechanistic interpretability? and Speaking to Congressional staffers about AI risk for understanding people’s takes on these areas.