TrikeApps has done a great job running LessWrong and adding new features, but they could use a little help. Have you thought about improving the LessWrong website but haven't done it because you weren't sure how? Or had installation issues? Well, now is a great time to start, because hacking on LessWrong just got much easier!
On behalf of the LessWrong Public Goods Team, I have built a Virtual Machine Image which hosts its own version of the LessWrong website. This eliminates the need to figure out how to host LessWrong yourself. To hack on LessWrong you simply:
- Install VirtualBox
- Download and use the VM image
- Edit LessWrong's code
- Test
- Submit pull request
Detailed instructions and download link here.
Interested, but not sure what to work on? The LessWrong issue tracker is here. Run into trouble with the code? Ask questions on the dev list.
Awesome!
I don't have time right now but someone should totally implement a recommendation engine for Less Wrong articles. Seems like it could be really high utility (is recency really the best criterion for determining the optimal Less Wrong article for a user to read?) as well as a great excuse to bone up on the latest in machine learning and narrow AI. (I'm a little leery of learning about a field without applying it to some project--if I'm able to apply knowledge successfully that makes me feel a lot more confident that I've mastered it. And having a project in mind helps resolve questions of how deeply I want to master a given piece of material--just master it well enough to do the project.) Plus if you fail at this there won't be any stigma--the reddit team worked on their recommendation engine for years without it going much of anywhere.
Can anyone recommend any books they've read that would be useful to an endeavor such as this? Programming Collective Intelligence looks pretty interesting but I haven't read it.
Better than recency, perhaps the top scoring posts of all time?
A recommendation engine needs information about what posts you are glad to have read and ideally what posts you read but did not fund useful. So if the engine knows for each user (1) the set of posts they've read [1] and (2) the set of posts that they've voted up, then we have an evaluation criterion: did we choose to show people posts they voted up?
You'd then need to figure out features and write code for them so the learning algorithm could find user correlations. Set up a svm or something ... (read more)