I’m currently researching forecasting and epistemics as part of the Quantified Uncertainty Research Institute.
ozziegooen(Ozzie Gooen)
Karma: 1,981
Just want to flag that I like the idea of topic-specific books, perhaps with an additional author to help rewrite things and make them consistent and clean. It’s especially enticing if you can find labor to do it that doesn’t have a high opportunity cost for other LessWrong style things.
+1 for being able to be open about small disagreements like this online :)
Thanks for the reasoning here. I also don’t want to detract people from purchasing these books, I imagine if people really wanted they could write the dates on them manually.
That said—
To better explain my intuitions here:
In 5 years from now, I care about whether the essays came out in 2018 or in 2017 if I am trying to find a particular one in a book, or recommend one to another person. Ordering is really simple to remember compared to other kinds of naming one could use. When going between different books the date is particularly relevant because names and concepts will change over time. I’d hope that 10 years from now much of the 2018 content will look antiquated and old.If you’re just aiming for “timeless and good quality posts” (this sounds like the value proposition for the readers you are referring to), then I don’t understand the need to only choose ones from 2018. Many good ones came out before 2018 that I imagine would be interesting to readers. That said, if you plan on releasing them on yearly intervals later I’d imagine some restriction might be necessary. Or, it could be that whenever a few topics seem to have come full circle or be in a good place for a book, you publish a book focused on those topics.
I agree that “LessWrong Review 2018” sounds strange, but there are other phrases that could have with 2018 in them. Many Academic periodicals (including things like Philosophy, which are at least as timeless as LessWrong content) have yearly collections. With those I don’t assume I need to read all of the old ones before reading the current year, that would take quite a while (it becomes more obvious after a few are out). I imagine the name could be something like, “LessWrong Highlighted Content: 2018″ or “The Best of LessWrong: 2018”.
It’s very possible that there’s kind of a “free pass” for the first 1-3 years, if this is a repeating thing, and then you could start adding the year. It’s not that big a deal if there are just 2-3 of these, but I imagine it will get to be annoying if there are 5+ (and by that time it will be more obvious if it’s an issue or not)
If the main thing that separates this book from the 2019 and 2020 books is that it’s the collection of posts from 2018, it’s counterintutive to me that that’s not the prominent feature of the title here. Other “journals of the year” often make the year really prominent.
I feel like 5 years from now I’m going to have trouble remembering that “A Map That Reflects the Territory” refers to the 2018 edition, and some other equally elegant but abstract name refers to the 2019 edition.
If you do go with really premium books especially, I’d recommend considering making the date the prominent bit. Honestly I expect to memorize the “lesswrong”ness from the branding (which is distinct), so the year seems like the most important part to me.
That said, I feel like I’m not exactly in the target audience (generally don’t prefer physical books), so it would come down to the preferences of others.
I realize you’ve probably thought about this a lot and have reasons, just giving my 2 cents.
The books look very pretty, nice work.
Is this content from 2018 specifically, or is it taken from all of historic LessWrong? My impression was that this was from the 2018 review, but I don’t see anything about that in the description above.
If it is from the 2018 review, do you have ideas on how you will differentiate the 2019/2020/etc versions?
Thank you for raising the issue. Happy to clarify further.
By evaluation we refer essentially[1] to the definition on Wikipedia page here.
Evaluation is a systematic determination of a subject’s merit, worth and significance, using criteria governed by a set of standards. It can assist an organization, program, design, project or any other intervention or initiative to assess any aim, realisable concept/proposal, or any alternative, to help in decision-making; or to ascertain the degree of achievement or value in regard to the aim and objectives and results of any such action that has been completed
By “interesting” we mean what will do well on the listed rubric. We’re looking for examples that would be informative for setting up new research evaluation setups. This doesn’t mean the examples have to deal with research, but rather that they bring something new to the table that could be translated. For example, maybe there’s a good story of a standardized evaluation that made a community or government significantly more or less effective.
[1] I say “essentially” because I can imagine that maybe someone will point out some unintended artifact in the definition that goes against our intuitions, but I think that this is rather unlikely to be a problem.
I find this interesting, thanks for working on it. I’ve been thinking about similar things for a while and have heard related discussions, but I’m happy to have more standardized terminology and the links to existing literature.
I am more interested in how this could be used improve our thinking abilities for broad range of valuable purposes, rather than on the implications specifically for them to be unsafe.
[Question] Prize: Interesting Examples of Evaluations
Thanks! Yea, this is quite similar to Guesstimate. I think and hope that in the future estimation/monte carlo tech will be closely integrated with forecasting systems.
Squiggle: Technical Overview
Thanks for the suggestion.
My background is more in engineering than probability, so have been educating myself on probability and probability related software for this. I’ve looked into copulas a small amount but wasn’t sure how tractable they would be. I’ll investigate further.
Thanks for the feedback!
It’s a good point. This is a kind of thing I’ve been wrestling a lot, though of course is fairly surface level compared to the main architecture.I don’t have a personal preference on this. I agree that mixture is more technically correct term, but it seems like many (not very technical people) find “multimodal” more intuitive. To many people I think “mixture” sounds more generic, as from where they are standing, “mixture” could mean several things.
I’ll keep the option in mind and ask for further preferences.
Squiggle: An Overview
I’m not sure what you are looking for. Most people know very little in the space of all the things one could find out in books and the like, much which is useful to some extent. If you’re curious what things I specifically think are true but the public doesn’t yet know of, then continue to read my blog posts; it’s a fair bit of stuff, but rather specific.
There are a few options with the $15/month package with Immersed. No forest, but there is one above the clouds, and one in a cave (no treasure though). With the free package you just get a few 360 photos to choose from (no depth)
Other apps have more options, but they only support Windows generally.
The default strap. It’s not that great, but for me, tolerable. I’m giving it a few months before upgrading, as I’m hoping more straps will be available. (the Oculus ones are sold out)
Fixed, thanks! It was a small error in how the url was typed.
Comparing groups of forecasters who worked on different question sets only using simple accuracy measures like brier scores is basically not feasible. You’re right that forecasters can prioritize easier questions and do other hacks.
This post goes into detail on several incentive problems:
https://forum.effectivealtruism.org/posts/ztmBA8v6KvGChxw92/incentive-problems-with-current-forecasting-competitions
I don’t get the impression that platforms like Metaculus or GJP bias their questions much to achieve higher brier scores. This is one reason why they typically focus more on their calibration graphs, and on direct question comparisons between platforms.All that said, I definitely think we have a lot of room to get better at doing comparisons of forecasting between platforms.
Kudos for the thinking here, I like the take.
There’s a whole lot to “making people more correct about things.” I’m personally a lot less focused on trying to make sure the “masses” believe things we already know, than I am in improving the epistemic abilities of “best” groups. From where I’m standing, I imagine even the “best” people have a long way to improve. I personally barely feel confident about a bunch of things and am looking for solutions where I could be more confident. More “super intense next level prediction markets” and less “fighting conspiracy theories”.
I do find the topic of epistemics of “the masses” to be interesting, it’s just different. CSER did some work in this area, and I also liked the podcast about Taiwan’s approach to it (treating lies using epidemic models, similar to how you mention.)
Fixed