So, twelve articles, one of them interesting, three or four have a good idea but are very long, and the rest feels useless.
I appreciate you took the time to read all of them (or enough to comment on them). I also feel some are better written than the others, and I was also more inspired for some. From what I understood, you want the articles to be “useful” and “not too long”. I understand what you would want that (maximize the (learned stuff)/(time spent on learning) ration). I used to write on Medium where the read ratio of posts would decrease significantly with the length of the post. This pushed me to read shorter and shorter posts, if I wanted to be read entirely. I wanted to try LW because I imagined here people would have longer attention spans and could focus on philosophical/mathematical thinking. However, if you’re saying I’m being “too long with very low density of ideas” I understand why this could be infuriating.
I typically do not downvote the “meh” articles, but that’s under assumptions that they don’t appear daily from the same author
I get your point, and it makes sense with what you said in the first comment. However, I don’t feel comfortable with people downvoting “meh” articles because of the author (even though it’s daily). I would prefer a website where people could rate articles independently of who the author is, and then check their other stuff.
My aggregate feedback would be: You have some good points. But sometimes you just write a wall of text.
Ok. So I should be more clear/concise/straight-to-the-point, gotcha.
And I suspect that the precommitment to post an article each day could be making this a lot worse. In a different situation, such as writing for an online magazine which wants to display a lot of ads, writing a lot of text with only a few ideas would be a good move; here it is a bad move.
Could you be more specific about what you think would be my move? For the online magazine, getting the maximum number of clicks/views to display the more ads makes sense, and so lots of text with lots of ads, and enough ideas to ensure the reader keeps seeing adds makes sense.
But what about LW? My move here was simple: understand better AI Safety by forcing myself to daily crystallize ideas about ideas related to the field, on a website with great feedback/discussions and low-tolerance for mistakes. For now, the result (in the discussions) is, overall, satisfying, and I feel that people here seem to enjoy AI Safety stuff.
More generally, I think the fact that if I generate 10% of headers or you get to click on all my articles may be correlated to other factors than me daily posting, such as:
The LW algorithm promotes them
You’re “Michaël Trazzi” filter (you need one, because you get to see my header) is not tuned correctly, because you still seem to still be reading them, even if only 1⁄12 felt useful (or maybe you just read them to comment on this post?).
This comment is already long (sorry for the wall of text), so I will say more about the Meta LW high/low quality debate on Elo’s comment below.
However, I don’t feel comfortable with people downvoting “meh” articles because of the author (even though it’s daily).
The “meh” articles should be downvoted. Simply because LW is the place where I come to read something better.
However, it makes sense strategically to be more lenient towards new authors. The reasoning is that it people are often scared to make the first post here, and that it takes some time to get attuned to the local culture. This there is a chance that a person who wrote a “meh” article first will write better articles later. On the other hand, having one’s only article downvoted is probably emotionally harder that having e.g. one of three articles downvoted. So it the new author’s first article is “meh”, instead of downvoting I often just abstain from voting and try to give some good advice in a comment.
While this may feel the same, the motivation is not punishing frequent authors (I would be quite happy with frequent high-quality articles), but rather making the first step into the community easier—with the tacit assumption that the author will improve.
I appreciate you took the time to read all of them (or enough to comment on them). I also feel some are better written than the others, and I was also more inspired for some. From what I understood, you want the articles to be “useful” and “not too long”. I understand what you would want that (maximize the (learned stuff)/(time spent on learning) ration). I used to write on Medium where the read ratio of posts would decrease significantly with the length of the post. This pushed me to read shorter and shorter posts, if I wanted to be read entirely. I wanted to try LW because I imagined here people would have longer attention spans and could focus on philosophical/mathematical thinking. However, if you’re saying I’m being “too long with very low density of ideas” I understand why this could be infuriating.
I get your point, and it makes sense with what you said in the first comment. However, I don’t feel comfortable with people downvoting “meh” articles because of the author (even though it’s daily). I would prefer a website where people could rate articles independently of who the author is, and then check their other stuff.
Ok. So I should be more clear/concise/straight-to-the-point, gotcha.
Could you be more specific about what you think would be my move? For the online magazine, getting the maximum number of clicks/views to display the more ads makes sense, and so lots of text with lots of ads, and enough ideas to ensure the reader keeps seeing adds makes sense.
But what about LW? My move here was simple: understand better AI Safety by forcing myself to daily crystallize ideas about ideas related to the field, on a website with great feedback/discussions and low-tolerance for mistakes. For now, the result (in the discussions) is, overall, satisfying, and I feel that people here seem to enjoy AI Safety stuff.
More generally, I think the fact that if I generate 10% of headers or you get to click on all my articles may be correlated to other factors than me daily posting, such as:
The LW algorithm promotes them
You’re “Michaël Trazzi” filter (you need one, because you get to see my header) is not tuned correctly, because you still seem to still be reading them, even if only 1⁄12 felt useful (or maybe you just read them to comment on this post?).
This comment is already long (sorry for the wall of text), so I will say more about the Meta LW high/low quality debate on Elo’s comment below.
The “meh” articles should be downvoted. Simply because LW is the place where I come to read something better.
However, it makes sense strategically to be more lenient towards new authors. The reasoning is that it people are often scared to make the first post here, and that it takes some time to get attuned to the local culture. This there is a chance that a person who wrote a “meh” article first will write better articles later. On the other hand, having one’s only article downvoted is probably emotionally harder that having e.g. one of three articles downvoted. So it the new author’s first article is “meh”, instead of downvoting I often just abstain from voting and try to give some good advice in a comment.
While this may feel the same, the motivation is not punishing frequent authors (I would be quite happy with frequent high-quality articles), but rather making the first step into the community easier—with the tacit assumption that the author will improve.