I’m thinking of thesepapers which were posted here only after they were finished and published. Also this one which I posted here because Carl didn’t. Also Paul Christiano posting stuff on his own blog instead of LW.
They may have improved the SNR, but I can’t say there seem to be very much feedback or ideas...
That’s strange. I find LW feedback useful on my posts, and assumed that would be the case for others. Can you give an example of a post that didn’t gather useful feedback and ideas?
In the first link, for three papers, there’s exactly one substantive comment on a paper
The second link has roughly 3 or 4 comment threads which revolve around a specific point which seemed to cause changes in the paper, with the rest of the comments being relatively unrelated.
The third link contains some interesting comments about the paper on a meta level, but nothing that could be useful to the author, IMO.
the power post’s few comments are dominated by citation format, matriarchy and why anyone cares. None of these were useful to me except maybe the format carping.
the Sobel post has maybe 2 or 3 comments of value
the intelligence failures link garnered 1 comment of value
I guess it wasn’t clear, but I was suggesting that if those papers had been posted here while they were still in draft form (as opposed to “finished and published”), they would have received more discussions since people would have more incentives to participate and potentially influence the final output.
As for your posts, I think the reason for lack of useful feedback is that they are mostly summaries of many academic papers and it’s hard to give useful feedback without spending a lot of time to read those papers which nobody has an sufficient incentive to.
I got some comments for mydrafts. There were some valuable suggestions in both threads which I incorporated, but I had hoped for a little more feedback.
If you post more drafts in the future, I think it would help to add more context: Who is the target audience? What are you hoping to accomplish with the papers? (If we knew that we might care more about helping you to improve them.) Do they contain any ideas that are new to LW?
I’m thinking of these papers which were posted here only after they were finished and published. Also this one which I posted here because Carl didn’t. Also Paul Christiano posting stuff on his own blog instead of LW.
That’s strange. I find LW feedback useful on my posts, and assumed that would be the case for others. Can you give an example of a post that didn’t gather useful feedback and ideas?
Well, look at your own links.
In the first link, for three papers, there’s exactly one substantive comment on a paper
The second link has roughly 3 or 4 comment threads which revolve around a specific point which seemed to cause changes in the paper, with the rest of the comments being relatively unrelated.
The third link contains some interesting comments about the paper on a meta level, but nothing that could be useful to the author, IMO.
As for my own feedback, I keep a public list in http://www.gwern.net/Links#fn2 Going backwards through the last 3:
the power post’s few comments are dominated by citation format, matriarchy and why anyone cares. None of these were useful to me except maybe the format carping.
the Sobel post has maybe 2 or 3 comments of value
the intelligence failures link garnered 1 comment of value
I guess it wasn’t clear, but I was suggesting that if those papers had been posted here while they were still in draft form (as opposed to “finished and published”), they would have received more discussions since people would have more incentives to participate and potentially influence the final output.
As for your posts, I think the reason for lack of useful feedback is that they are mostly summaries of many academic papers and it’s hard to give useful feedback without spending a lot of time to read those papers which nobody has an sufficient incentive to.
I got some comments for my drafts. There were some valuable suggestions in both threads which I incorporated, but I had hoped for a little more feedback.
If you post more drafts in the future, I think it would help to add more context: Who is the target audience? What are you hoping to accomplish with the papers? (If we knew that we might care more about helping you to improve them.) Do they contain any ideas that are new to LW?
Thanks, that’s a good suggestion.