Coding day in and out on LessWrong 2.0
Also, this is great, and I am sad I didn’t know about this for the last month!
I… think this post was impacted by a bug in the LW API that GreaterWrong ran into, that made it so that it wasn’t visible on the frontpage when it was published. It nevertheless appears to have gotten some amount of engagement, but maybe that was all from direct links?
Given the substantial chance that a number of people have never seen this post, I reposted it. Its original publishing date was the 11th of June.
Note: Due to a bug in the LW API that GreaterWrong ran into this post was not visible on the frontpage until now. I reset it’s date to make sure that it gets any time in the frontpage (even if it’s a month after it’s original publishing date, which was the 12th of September)
Note: Due to a bug in the LW API that GreaterWrong ran into, this post wasn’t visible until now, but was published 3 days ago. I reset its date to give it proper time on the frontpage.
Ok, I now updated all the images in the post to be properly resized.
Yeah, some of the images are really really massive. I will reupload some more of them with smaller resolutions.
The first one was really important because it was shipped with every frontpage load, so I didn’t notice the second one.
We already compress images if you upload them to LW, but this post was cross posted, where we just take the HTML and copy it over, and we don’t have any fancy infrastructure set up to reupload and resize the images for that use case.
Multicore gained some favor with me when he did an enormous amount of tagging during the tagging sprint. Figured I would use my entry for the good, even if I didn’t have time to write my own thing.
Mod note: the first image in the post was 27000px wide, making it over 10MB large. I reduced it’s size to safe the poor data plans of our mobile connections and to make the post (and the frontpage) load a bit faster.
Alas, looks like Scott’s old blog that was hosted on raikoth.net is now captured by some random malware distributor. I fixed the link on that specific post with a repost version, but there are probably dozens of other links to that URL all over LessWrong.
Promoted to curated: This really is a quite good and straightforward explanation of what I think of as one of the really core ideas in theoretical rationality, covering surprisingly much ground in a way that feels approachable. I remember how impactful reading “A technical explanation of a technical explanation” was for my thinking about rationality, and this feels like a much less daunting introduction to some of the same ideas.
Really sorry for the downtime for parts of the last hour. The cause was a straightforward merge conflict that wasn’t highlighted by Git, and in a particularly sad instance of carelessness we merged even though our tests told us the build wasn’t working. This was basically fully a process failure on our side to not pay attention to the systems we put in place to prevent exactly this happening, and pressing big red override buttons because a previous problem with our CI-system had forced us to press some override buttons frequently enough that we weren’t thinking of it as something to be really careful with. Again, sorry for that.
I already made some of the most obvious changes to our process to prevent this from happening again, and am working on a bunch of larger changes that will make stuff in this reference class much less bad.
I know two people who I think trialed for it, but it never really worked out. But not confident about why, or whether MIRI is still interested in filling that position.
Hmm, definitely feels core to the art of rationality to me. Like, convergent instrumental goals apply to humans as well. Understanding that just feels straightforwardly useful for the generalized art of rationality.
Alas, these ones don’t work for me.
Man, this one is so great. I want to have a statue like this in my garden now.
Scope sensitivity and the cosmic endowment. I definitely feel like looking at the stars reminds me of how much stuff there is to optimize, which sure feels pretty related to rationality.
Huh, never experienced anything like this. CMD+Enter (on Mac) or CTRL+Enter (on Windows) has always submitted a comment (and does so on basically any web-platform that I know of)
I try and post something starting with a copypasted quote. How do I get out of quote mode?
You get out of quote mode by pressing enter twice. Which I think is standard in most text-editors. Alternatively you can also select any paragraph in the blockquote and press the blockquote button.
I want to split a paragraph. How? No idea. Nothing seems to work.
You can split a paragraph by pressing enter with your cursor at the place you want to split it, which really seems like the obvious thing. What happens when you press enter?
Is there any chance your CMD/CTRL key is broken? The only way I can think of to explain what’s happening here is to hypothesize that you must be pressing CMD+Enter or CTRL+Enter a lot, even though there should basically never be a reason to use those combinations (Shift+Enter allows you to insert a manual line-break without a paragraph break, which is something that some other platforms do with CTRL+Enter, at least on Mac, so maybe that’s what’s confusing you?)
Great! For people who are bothered by the brightness, this seems like a decent solution for now. We are thinking about creating styling for a proper dark mode, but really unclear what the timeline on that is atm.
The model of “sacrifices to the gods” is a very specific gears-level model of how decisions get made. It’s a relatively straightforward application of Hanson’s theory of costly signaling, and as such has a lot of content to it that can be proven or disproven.
I think it’s really important to distinguish arguments that are tribal and act on an associative level, which I think are generally bad, from arguments that operate on a gears-level and make concrete predictions, even if those predictions are uncharitable. The first one does often seem bad and to deteriorate discussion, but the second one seems quite important and usually moves discussion forward. It seems to me that Zvi’s use of “sacrifices to the gods” falls into the second camp here.