I wanted a datapoint for Czynski’s hypothesis that LW 2.0 killed the comment sections, so I checked how many comments your blogposts were getting in the first 3 months of 2017 (before LW 2.0 rebooted). There were 13 posts, and the comment counts were 0, 0, 2, 6, 9, 36, 0, 5, 0, 2, 0, 0, 2. (The 36 was a a political post in response to the US election, discussion of which I generally count as neutral or negative on LW, so I’d discount this.)
I’ll try the same for Zvi. 13, 8, 3, 1, 3, 18, 2, 19, 2, 2, 2, 5, 3, 7, 7, 12, 4, 2, 61, 31, 79. That’s more active (the end was his excellent sequence Against Facebook, and the last one was a call for people to share links to their blogs).
So that’s not zero, there was something to kill. How do those numbers compare during LessWrong 2.0? My sense is that there’s two Zvi eras, there’s the timeless content (e.g. Mazes, Sabbaths, Simulacra) and the timeful content (e.g. Covid, AI, other news). The latter is a newer, more frequent, less deep writing style, so it’s less apples to apples, so instead let’s take the Moral Mazes sequence from 2020 (when LW 2.0 would’ve had a lot of time to kill Zvi’s comments). I’m taking the 17 posts in this main sequence and counting the number of comments on LW and Wordpress.
#
LW
Wordpress
1
16
5
2
40
19
3
29
23
4
8
12
5
7
21
6
56
10
7
6
13
8
12
8
9
18
8
10
21
18
11
26
21
12
42
16
13
6
11
14
9
15
15
14
18
16
11
19
17
28
22
SUM
349
259
This shows the comment section on Wordpress about as active as it was in the 3-month period above (259 vs 284 comments) in the 2-months that the Mazes sequence was released, and comments were more evenly distributed (median of 17 vs 5). And it shows that the LessWrong comment section more than doubled the amount of discussion of the posts, without reducing the total discussion on Zvi’s wordpress blog.
These bits of data aren’t consistent with LW killing other blogs. FWIW my alternative hypothesis is that these things are synergistic (e.g. I also believe that the existence of LessWrong and the EA Forum increases discussion on each), and I think that is more consistent with the Zvi commenting numbers.
Comments on my own blog are almost non existent, all the interesting discussion happens on LW and Twitter.
(Full disclosure: am technically on mod team and have deep social ties to the core team)
I wanted a datapoint for Czynski’s hypothesis that LW 2.0 killed the comment sections, so I checked how many comments your blogposts were getting in the first 3 months of 2017 (before LW 2.0 rebooted). There were 13 posts, and the comment counts were 0, 0, 2, 6, 9, 36, 0, 5, 0, 2, 0, 0, 2. (The 36 was a a political post in response to the US election, discussion of which I generally count as neutral or negative on LW, so I’d discount this.)
I’ll try the same for Zvi. 13, 8, 3, 1, 3, 18, 2, 19, 2, 2, 2, 5, 3, 7, 7, 12, 4, 2, 61, 31, 79. That’s more active (the end was his excellent sequence Against Facebook, and the last one was a call for people to share links to their blogs).
So that’s not zero, there was something to kill. How do those numbers compare during LessWrong 2.0? My sense is that there’s two Zvi eras, there’s the timeless content (e.g. Mazes, Sabbaths, Simulacra) and the timeful content (e.g. Covid, AI, other news). The latter is a newer, more frequent, less deep writing style, so it’s less apples to apples, so instead let’s take the Moral Mazes sequence from 2020 (when LW 2.0 would’ve had a lot of time to kill Zvi’s comments). I’m taking the 17 posts in this main sequence and counting the number of comments on LW and Wordpress.
This shows the comment section on Wordpress about as active as it was in the 3-month period above (259 vs 284 comments) in the 2-months that the Mazes sequence was released, and comments were more evenly distributed (median of 17 vs 5). And it shows that the LessWrong comment section more than doubled the amount of discussion of the posts, without reducing the total discussion on Zvi’s wordpress blog.
These bits of data aren’t consistent with LW killing other blogs. FWIW my alternative hypothesis is that these things are synergistic (e.g. I also believe that the existence of LessWrong and the EA Forum increases discussion on each), and I think that is more consistent with the Zvi commenting numbers.
I was part of the 2.0 reboot beta: there are no posts of mine on LW before that
I still prefer the ones I see there to what I see on LW. Lower quantity higher value.