“—Will you always remain a philosopher? His mother once timidly asked him.
-And what is a philosopher? He turned around.
-A madman who tortures himself all his life so that people will talk about him after death.”
Oblivion Twist
Karma: 0
“—Will you always remain a philosopher? His mother once timidly asked him.
-And what is a philosopher? He turned around.
-A madman who tortures himself all his life so that people will talk about him after death.”
Apology for possible duplication
I am sure there have been numerous similar posts on the forum before. However I have not found any points on my specific idea. I do apologize in advance if I missed the relevant thread.
Background: My rejected post
As a new user to LessWrong forum I made a post that was rejected as LLM written. I did specify in the header of the post that I did use LLM for editing purposes since English is not my first language and I often have doubts about my phrasing and articulation even in my native language, but the ideas and concepts were mine.
I find it ironic that LessWrong managers seek genuine human input but use automated LLM-detection services and posts get rejected without ever being reviewed by a human. As a rationalist I totally understand the managers of the forum they do not have human resources and capacity to manually check every post.
My proposed experiment
I do not want to impose my ideas on how to improve the situation. It is not my forum to improve. However I had a speculative thought how I might help the managers.
Plan: I’ve decided that once a day/week/month I’ll read a rejected post and make a comment on it. It does not matter if I like it or not. It does not matter if I agree with it or not but just so the new user who got their post rejected will see that at least one real human has read it.
In that comment I would encourage the author of the rejected post to do the same to another rejected post that does not have any comments under it yet. If the numbers grow and let’s say 100 new authors are checking one novel idea on regular basis that might make a difference.
Caveats:
Maybe it is sensible to restrict the number of comments to rejected posts so that it does not divide the forum and grow a new community of “rejects”.
I do not propose to comment on every rejected post. It will be physically impossible. In addition I believe in the good faith and common sense of newcomers. Spam will always be obvious spam. Trolling will always stay obvious trolling.
My main point is to support newcomers and not get them discouraged. (I already have an interesting rejected post in mind by a young user (self-identified as 15). It is frustrating when your post gets rejected at any age but for a teenager it could be very discouraging.)
Self-reflection from manager’s perspective
I have thought about the idea as critically as I could. I put myself in the shoes of LessWrong manager. I have been running this forum for years. I have addressed multiple issues with numerous approaches and managed to achieve something coherent. Here is some newcomer who believes they know better. As a human I would get defensive and most likely reject the idea without even reading it thoroughly.
On the other hand if my post did not get rejected I would never write this particular comment. So maybe the karma system is already operating at optimal settings given the circumstances and my entire hypothesis is false.
Questions
I wonder if it would be ethical to attempt such experiment? Would it break any rules of LessWrong? Is it technically possible/hard to implement?
P.S. I have whole other essay on why I use LLM text polishing.
P.P.S. This comment is not an attempt to change the status of my rejected post. In fact now I believe it was: 1. Naïve. 2. Not fully aligned with the spirit of the forum.