I’m not very worried about an endless September. LW is pretty good at downvoting people when they make rookie mistakes in reasoning and argument, or when they are mean or trollish. The new troll toll (or whatever solution we settle into after a few months) should go even further toward preventing endless September. Moreover, I think the content itself here does a fairly good job of filtering out many kinds of people we don’t want.
Finally, I think Xachariah’s point is important: “the eternal September effect is primarily caused by new-member with new-member interaction.” I would say that LW already does a good job of limiting this. For example, new members who don’t understand the culture are downvoted, which means their comments are hidden by default. Also, people are already incentivized to lurk for quite a while before commenting or posting, because the community is clearly intelligent and is constantly using community jargon they could easily be downvoted for misunderstanding.
I also don’t think we should make it harder for people to join (e.g. with a quiz). Instead, I think we should make it even easier for the kinds of people we want to find LW and engage. Here are some ideas for doing that:
Improve the wiki (already in progress).
Create a nicely-formatted ebook of The Sequences (already in progress).
Have an army of volunteers regularly comment on selected blogs and discussion forums (e.g. for computer scientists, cognitive scientists, mathematicians, and formal philosophers), linking back to the stickest relevant LW posts. (This project is under development: if you want to be part of this once it’s ready, please notify malo@intelligence.org.)
Create an attractive “Welcome to Less Wrong” page that can quickly guide people to posts they’re most likely to be interested in. This would include a video of somebody explaining in 5-10 minutes what Less Wrong is about, the values of the community, etc. (Project under development.)
Create a “community values” page. I like this idea of yours very much. The values could be pretty simple, things like: (1) Don’t be a dick. “Do unto others 20% better than you expect them to do unto you, to correct for subjective error.” (2) When disagreeing, aim for the higher levels of the disagreement hierarchy. (3) Be clear. Be specific. Give examples. Hyperlink to longer explanations or relevant articles. (4) Before commenting, try to list some reasons you might be wrong about the thing you’re about to say. (5)...
Have the system deliver a welcome message to a user’s inbox when they first sign up, one linking to the “community values” page, the “welcome to LW” page, etc.
Create handy one-stop FAQs (on rationality, decision theory, etc.) that become standard resources for their subjects on the internet, and make sure they link heavily to LW. (This project is also in progress.)
I don’t think I like the idea of splitting the community into hard and easy forums, nor the idea of inducting something like Crocker’s rules into the code of conduct.
This would include a video of somebody explaining in 5-10 minutes what Less Wrong is about, the values of the community, etc. (Project under development.)
What’s the case for a video? Seems a little cheesy, IMO.
Have an army of volunteers regularly comment on selected blogs and discussion forums (e.g. for computer scientists, cognitive scientists, mathematicians, and formal philosophers), linking back to the stickest relevant LW posts. (This project is under development: if you want to be part of this once it’s ready, please notify malo@intelligence.org.)
OK, but let’s make sure they really do have some domain expertise in the area that they’re leaving comments in, so they don’t make us look bad. Link.
Have the system deliver a welcome message to a user’s inbox when they first sign up, one linking to the “community values” page, the “welcome to LW” page, etc.
I like this idea. One potential logistical glitch: If the user isn’t already familiar with reddit, they won’t know what an orange envelope means and they may just see it as orange forever and never click on it.
People like videos? I hate videos to the point that I will go out of my way to avoid links with videos in them, and I’ve seen this sentiment expressed by other people here.
I hate video because it goes too slow. I can read at least twice as fast as a video goes. It always feels like such an excruciating waste of time. Also, I can’t use find in page. I am addicted to find in page. Ctrl-F and me are attached at the hip. Of all the pages I open, the proportion I read in entirety is very small. Ctrl-F is like half my way of navigating the internet. I’m really glad to see someone else express this. I thought i was the only one.
I like videos. They are more passive than written text and feel less cognitively demanding per unit time. In fact, I will often prefer to watch/listen-to a video/audio recording more than once in order to achieve the same level of retention as reading text in a concentrated fashion, thereby exchanging time for concentration-willpower.
Transcripts are fairly expensive; patio11 pays for transcripts to be made for his podcasts (a big factor in why those submissions do well on Hacker News), but IIRC the quoted figure is north of $100. So you would pay… but would you pay enough?
Echoing komponisto, my job is incredibly non demanding of my cognitive resources so I constantly listening to audiobooks, youtube channels, and TCC/TMS Lectures at 2x speed. Over the course of an 8 hr work day I can finish about 200 pages at reasonable comprehension.
I’d be curious to hear your evidence for this. In any case, even if there is conclusive evidence that internet users prefer video presentations over corresponding text presentations, it’s not obvious that this trend extends to LWers or potential LWers.
Also, this seems to have been a flop. I suspect that if videos were a good fit for LW concept transmission, we’d have seen more success with that small experimental effort.
Those video experiments were very poorly produced. That’s not the kind of video I have in mind. And video would of course only be there in addition to text.
I would have enjoyed and reccomended even poorly produced videos if you guys had bothered to extend them. I keep meaning to finish the last third or so of the sequences I haven’t read, but their all over the place and it makes sense for me to start from the top. It’d be great if I could listen while doing other things. In my case, painting mostly, in other cases, probably cleaning, laundry, dishes, pet care and other activities that take up very low or no verbal mental resources.
Guys. It’s not rocket science. You’re smart. You have good content. Present it well. Or better. If you can’t do that, hire someone who can. Get it out there. If you can’t do that, hire someone who can.
You are the decision maker here who will determine whether any changes go into effect. If this is your final decision on the matter of preventing endless September, let me know, and I will place a note at the top of this discussion to prevent people from continuing to waste time on it. If not, then I will focus on debating it with you because I still disagree but it would be a waste of time for me to move forward if you feel that LessWrong needs no more protection against endless September.
Okay. I will start a new post specifically as a call to agreement for us decide whether LessWrong should have better endless September protection. It will take me a little while to get it all organized. Give me a bit of time.
Well I thought it out carefully and added in citations and whatnot, so now it’s kinda too long for a comment. Sorry you did not get the post where you wanted it, but it’s done now. here it is.
I’m not very worried about an endless September. LW is pretty good at downvoting people when they make rookie mistakes in reasoning and argument, or when they are mean or trollish. The new troll toll (or whatever solution we settle into after a few months) should go even further toward preventing endless September. Moreover, I think the content itself here does a fairly good job of filtering out many kinds of people we don’t want.
Finally, I think Xachariah’s point is important: “the eternal September effect is primarily caused by new-member with new-member interaction.” I would say that LW already does a good job of limiting this. For example, new members who don’t understand the culture are downvoted, which means their comments are hidden by default. Also, people are already incentivized to lurk for quite a while before commenting or posting, because the community is clearly intelligent and is constantly using community jargon they could easily be downvoted for misunderstanding.
I also don’t think we should make it harder for people to join (e.g. with a quiz). Instead, I think we should make it even easier for the kinds of people we want to find LW and engage. Here are some ideas for doing that:
Improve the wiki (already in progress).
Create a nicely-formatted ebook of The Sequences (already in progress).
Have an army of volunteers regularly comment on selected blogs and discussion forums (e.g. for computer scientists, cognitive scientists, mathematicians, and formal philosophers), linking back to the stickest relevant LW posts. (This project is under development: if you want to be part of this once it’s ready, please notify malo@intelligence.org.)
Create an attractive “Welcome to Less Wrong” page that can quickly guide people to posts they’re most likely to be interested in. This would include a video of somebody explaining in 5-10 minutes what Less Wrong is about, the values of the community, etc. (Project under development.)
Create a “community values” page. I like this idea of yours very much. The values could be pretty simple, things like: (1) Don’t be a dick. “Do unto others 20% better than you expect them to do unto you, to correct for subjective error.” (2) When disagreeing, aim for the higher levels of the disagreement hierarchy. (3) Be clear. Be specific. Give examples. Hyperlink to longer explanations or relevant articles. (4) Before commenting, try to list some reasons you might be wrong about the thing you’re about to say. (5)...
Have the system deliver a welcome message to a user’s inbox when they first sign up, one linking to the “community values” page, the “welcome to LW” page, etc.
Create handy one-stop FAQs (on rationality, decision theory, etc.) that become standard resources for their subjects on the internet, and make sure they link heavily to LW. (This project is also in progress.)
I don’t think I like the idea of splitting the community into hard and easy forums, nor the idea of inducting something like Crocker’s rules into the code of conduct.
What’s the case for a video? Seems a little cheesy, IMO.
OK, but let’s make sure they really do have some domain expertise in the area that they’re leaving comments in, so they don’t make us look bad. Link.
I like this idea. One potential logistical glitch: If the user isn’t already familiar with reddit, they won’t know what an orange envelope means and they may just see it as orange forever and never click on it.
People like videos and it makes the community more human to newcomers.
People like videos? I hate videos to the point that I will go out of my way to avoid links with videos in them, and I’ve seen this sentiment expressed by other people here.
I hate video because it goes too slow. I can read at least twice as fast as a video goes. It always feels like such an excruciating waste of time. Also, I can’t use find in page. I am addicted to find in page. Ctrl-F and me are attached at the hip. Of all the pages I open, the proportion I read in entirety is very small. Ctrl-F is like half my way of navigating the internet. I’m really glad to see someone else express this. I thought i was the only one.
I like videos. They are more passive than written text and feel less cognitively demanding per unit time. In fact, I will often prefer to watch/listen-to a video/audio recording more than once in order to achieve the same level of retention as reading text in a concentrated fashion, thereby exchanging time for concentration-willpower.
I suppose I have nothing to complain about as long as the transcript is present and easy to get to.
Some do and some don’t.
I seem to recall lots of complaints on lukeprog’s first Q&A about the fact that the answers were delivered in video format.
FYI he also provided a text transcript.
[comment deleted]
Transcripts are fairly expensive; patio11 pays for transcripts to be made for his podcasts (a big factor in why those submissions do well on Hacker News), but IIRC the quoted figure is north of $100. So you would pay… but would you pay enough?
[comment deleted]
College students would be flaky and unreliable, and you’d want at least 2 for error-checking. You get what you pay for.
Confirmation bias and selection effects?
Echoing komponisto, my job is incredibly non demanding of my cognitive resources so I constantly listening to audiobooks, youtube channels, and TCC/TMS Lectures at 2x speed. Over the course of an 8 hr work day I can finish about 200 pages at reasonable comprehension.
I’d be curious to hear your evidence for this. In any case, even if there is conclusive evidence that internet users prefer video presentations over corresponding text presentations, it’s not obvious that this trend extends to LWers or potential LWers.
Also, this seems to have been a flop. I suspect that if videos were a good fit for LW concept transmission, we’d have seen more success with that small experimental effort.
Those video experiments were very poorly produced. That’s not the kind of video I have in mind. And video would of course only be there in addition to text.
I would have enjoyed and reccomended even poorly produced videos if you guys had bothered to extend them. I keep meaning to finish the last third or so of the sequences I haven’t read, but their all over the place and it makes sense for me to start from the top. It’d be great if I could listen while doing other things. In my case, painting mostly, in other cases, probably cleaning, laundry, dishes, pet care and other activities that take up very low or no verbal mental resources.
Guys. It’s not rocket science. You’re smart. You have good content. Present it well. Or better. If you can’t do that, hire someone who can. Get it out there. If you can’t do that, hire someone who can.
You are the decision maker here who will determine whether any changes go into effect. If this is your final decision on the matter of preventing endless September, let me know, and I will place a note at the top of this discussion to prevent people from continuing to waste time on it. If not, then I will focus on debating it with you because I still disagree but it would be a waste of time for me to move forward if you feel that LessWrong needs no more protection against endless September.
No, I want to be debated. I might change my mind.
Okay. I will start a new post specifically as a call to agreement for us decide whether LessWrong should have better endless September protection. It will take me a little while to get it all organized. Give me a bit of time.
Or, we could debate the subject here.
Well, you were right. Thumb up.
Well I thought it out carefully and added in citations and whatnot, so now it’s kinda too long for a comment. Sorry you did not get the post where you wanted it, but it’s done now. here it is.