{"items": [{"author": "Jeff&nbsp;Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639680592842", "anchor": "fb-639680592842", "service": "fb", "text": "@Ben", "timestamp": "1386021055"}, {"author": "Danner", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639684405202", "anchor": "fb-639684405202", "service": "fb", "text": "Heh, I thought you meant Electronic Arts...and was like \"you mean they ever had quality?\"", "timestamp": "1386023594"}, {"author": "Peter", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639684934142", "anchor": "fb-639684934142", "service": "fb", "text": "An interesting test would be to randomly subsample some posts and see if you can identify what time period it comes from based on the quality.  One would have to make sure that the posts can't be identified by context, though (i.e., explicitly referencing a historical event, etc.).", "timestamp": "1386023882"}, {"author": "Jeff&nbsp;Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639689115762", "anchor": "fb-639689115762", "service": "fb", "text": "@Peter: if you're interested in doing that the machine readable dump is here: http://www.jefftk.com/ea-fb-2013-12-02.json<br><br>I considered it, but I figured most people in a position to evaluate the quality of a post would also remember the specific posts.", "timestamp": "1386025797"}, {"author": "Jordy", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639689854282", "anchor": "fb-639689854282", "service": "fb", "text": "This is what I thought of http://www.youtube.com/watch?v=_tfup-Ztb4g", "timestamp": "1386026237"}, {"author": "David&nbsp;Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639695323322", "anchor": "fb-639695323322", "service": "fb", "text": "If I had an impression that the quality was changing, it would probably be more about the comments than the posts. (Ignoring useless posts is pretty easy. But having a good discussion with a bunch of noisy comments is hard.) So I thought I would choose a random sample of posts and count the number of comments and how many I thought were valuable, and then plot the ratio (or something) against time.<br><br>Then I realized this is probably a pretty silly way to spend my time, and I wouldn't be able to transparently share the results without offending people.", "timestamp": "1386029476"}, {"author": "Peter", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639695532902", "anchor": "fb-639695532902", "service": "fb", "text": "David&nbsp;Chudzicki: If you report the data in the aggregate, it shouldn't offend anyone.  I was thinking of creating a little mini app that allowed people to rate discussion quality of random discussions, but I'm not really sure this is worth my time.  I agree that needing comments is important for judging quality, though.", "timestamp": "1386029581"}, {"author": "David&nbsp;Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639696271422", "anchor": "fb-639696271422", "service": "fb", "text": "Not worth it for me. :) But in case anyone's interested, here's the script (edit: in the R programming language) I wrote to pick a few random threads and then load them in my browser:<br><br>install.packages(\"rjson\")<br>library(\"rjson\")<br>eaPosts &lt;- fromJSON(readLines(\"http://www.jefftk.com/ea-fb-2013-12-02.json\"))<br>eaPosts[length(eaPosts)]<br>set.seed(897897)<br>samples &lt;- sample.int(length(json_data), 10)<br>for (i in samples) browseURL(paste0(\"http://www.facebook.com\", eaPosts[[i]][[2]]))", "timestamp": "1386029940"}, {"author": "Patrick", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639698696562", "anchor": "fb-639698696562", "service": "fb", "text": "Ok, this is meta-discussion oneupsmanship lol. Jeff, did you seriously just post a discussion topic about the state of other discussions topics in the aggregate? And the original topics in discussion are themselves about the discourse surrounding philanthropy. There is an intellectual trope of the form \"now let's take a minute look at this issue from a broader perspective.\"  This feels like taking that device one iteration amusingly too far haha.", "timestamp": "1386031483"}, {"author": "Patrick", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639699240472", "anchor": "fb-639699240472", "service": "fb", "text": "Oh, I missed a level lol. Jeff is posting a topic for discussion about a topic someone else posted about topics that other people have posted about the discourse surrounding EA. That's like 5 level deep meta.", "timestamp": "1386031792"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639702528882", "anchor": "fb-639702528882", "service": "fb", "text": "Jeff&nbsp;Kaufman, I'm getting a blank page?", "timestamp": "1386033517"}, {"author": "Jeff&nbsp;Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639702613712", "anchor": "fb-639702613712", "service": "fb", "text": "@Ben: a blank page at http://www.jefftk.com/ea-group-history ?", "timestamp": "1386033591"}, {"author": "Jeff&nbsp;Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639702643652", "anchor": "fb-639702643652", "service": "fb", "text": "Which browser?", "timestamp": "1386033603"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639703047842", "anchor": "fb-639703047842", "service": "fb", "text": "Oh, it looks like it might just have been taking a while. Chrome on Linux.", "timestamp": "1386033653"}, {"author": "Ben", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639703292352", "anchor": "fb-639703292352", "service": "fb", "text": "Okay, it loaded once but not the other few times.", "timestamp": "1386033751"}, {"author": "Jeff&nbsp;Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=639703786362", "anchor": "fb-639703786362", "service": "fb", "text": "@Ben: it's doing a lot of js.  This is probably related to why fb makes you load incrementally.", "timestamp": "1386034029"}, {"author": "Jay", "source_link": "https://www.facebook.com/jefftk/posts/639680562902?comment_id=640782958692", "anchor": "fb-640782958692", "service": "fb", "text": "The noise is great.  Regardless of dumb comments (mine or  others' :) interspersed among the many very intelligent ones, I've learned tons about how to be a more effective altruist from the Fb group, and so have many of the friends I have added to it.  Deeper, more serious, snobby discussions should head for archivable, moderatable venues such as effective-altruism.org.  As well as academic journals.  The Fb group is great for what it is: a common room for the quick exchange of news and ideas.", "timestamp": "1386727792"}]}