One of my first thoughts was glosses.

If I recall, in the early middle ages, one of the main ways by which philosophy and proto-science advanced was through the extensive use of glosses. (as adapted from biblical glosses) Contemporary thinkers would all write commentaries on various works of Aristotle. At first, these were confined to the margins of the manuscripts being copied, but later they were published separately.

Since Aristotle had a fairly comprehensive philosophy, this meant reading all the glosses on a particular work of his brought you up to speed with the current state of knowledge on that topic. This had the effect of creating domains of knowledge, and scholarly specialization first nucleated around individual texts.

I say this, because one of the main problems with science today is just that there is so much of it. This makes it difficult to have interdisciplinary exchange of knowledge and meaningful communication and coordination.

Having a search engine like Google Scholar helps enormously, but it can be difficult to sift through a body of knowledge if you don't already know all the right keywords to search for. The existence of review articles also helps summarize, but it's still a somewhat clumsy solution. Why not replace the review article with a wiki?

It would be nice to have everything arranged formally and hierarchically, by field, sub-field, and then by topic within that sub field. Each level could have their own publicly editable summary, if there was enough human effort to maintain it. Imagine all that also ordered by citation index, and with links to all relevant news articles, blogs, and reddit threads commenting on each article.

Read a tabloid headline starting with "Scientists Say..."? Go directly to the wiki, and check what other scientists and the internet think of the research quality, background context, reputability, etc. Maybe even have a prediction market on whether the findings will replicate.

A huge part of the scientific discourse is no longer happening in the journal articles themselves, but this could capture it all in one place.

Depends how much storage space you are willing to buy.

One of my fantasies is a Raspberry Pi that automatically downloads all Wikipedia updates each month or so, to keep a local copy. The ultimate version of this would do the same for every new academic article available on Sci-Hub.

Sci-Hub is the largest collection of scientific papers on the planet, and has over 58 million academic papers. If they average 100 kB a piece, that's only 5.8 TB. If they average 1MB each, then you would need to shell out some decent cash, but you could in theory download all available academic papers.

Someone may even have already done something like this, and put the script on GitHub or somewhere. (I haven't looked.)

(Also, nice username. :) )

EDIT: It turns out there's a custom built app for downloading and viewing Wikipedia in various languages. It's available on PCs, Android phones, and there's already a version made specially for the Pi: http://xowa.org/home/wiki/Help/Download_XOWA.html

I wonder how difficult it would be to translate all of Sci-Hub into a wiki format that the app could add and read. You'd probably have to modify the app slightly, in order to divide up all the Sci-Hub articles among multiple hard drives. It might make the in-app search feature take forever, for instance. And obviously it wouldn't work for the Android app, since there's not enough space on a MicroSD card. (Although, maybe a smaller version could be made, containing only the top 32GB of journal articles with the most citations, plus all review articles.)

Even just converting science into a Wikipedia-like format would be useful for the sake of open access. Imagine if all citations in a paper were a hyperlink away, and the abstract would display if you hovered your mouse over the link. (The XOWA app does this for Wikipedia links.)

If you are looking for a sense of wonder in particular, I'd recommend pretty much anything by Carl Sagan. Audio-books, Cosmos show, etc. My thought is that the social proof will transfer his positive valence to you. Ideally, when thinking about anything on a grand scale, it would trigger an association with that sense of wonder.

I hesitated to say this though, since it is also likely to make a sense of smallness more salient, if it isn't already salient. Use your best judgement.

I'm not sure how much this would help you in particular, but spaced repetition, when done right, should jog your memory and make you work to recall something just before you would have forgotten it.

In order to learn and remember to apply useful concepts, I have an Anki deck containing the following:

• http://mcntyr.com/52-concepts-cognitive-toolkit/

• https://medium.com/@yegg/mental-models-i-find-repeatedly-useful-936f1cc405d#.qtvgobrvk

• The LW Wiki

• Miscellaneous, domain-specific tools and knowledge

For perspective, I find Elo's writing interesting maybe half the time. That's about on par with a random LW post, for me. (Whereas, >99% of facebook posts are uninteresting.)

If he published more than about once a day, or put a little less effort into each post, I think he'd lower the LW average. (According to my subjective judgement.) Conversely, another hour or so on each post, or a slightly higher self-filter might raise the average. (Assuming his idea of what makes a good post is fairly representative of mine.)

EDIT: Perhaps I should say why this is relevant. Xrisk isn't just things which could destroy humanity outright, but also things from which we never recover. I'm also interested in building robust institutions which can survive unexpected circumstances, and have positive impact over centuries to influence the far future. (Nobel Prize Foundaton, DARPA's 100 year Starship, Long Now Foundation, etc.) Perhaps cryonicists will also find it interesting.

There was a recent TED talk on what makes systems robust in changing environments.

He starts with the example of the immune system, then mentions long-lived social systems (catholic church, roman empire), but goes on to focus mainly on applications to businesses that want to survive black swan events and industry disruption.

This flows somewhat counter to conventional wisdom, which says to optimize for growth by putting all your eggs into whichever basket has the largest growth rate or market.

He lists 6 characteristic that all these robust, long-lived systems have in common: Redundancy, Diversity, Modularity, Adaptation, Prudence, and Embeddedness.

This is a good point. Pretty much all the things we're optimizing for which aren't our values are due to coordination problems. (There's also Akrasia/addiction sorts of things, but that's optimizing for values which we don't endorse upon reflection, and so arguably isn't as bad as optimizing for a random part of value-space.)

So, Moloch might optimize for things like GDP instead of Gross National Happiness, and individuals might throw a thousand starving orphans under the bus for a slightly bigger yacht or whatever, but neither is fully detached from human values. Even if U(orphans)>>U(yacht), at least there’s an awesome yacht to counterbalance the mountain of suck.

I guess the question is precisely how diverse human values are in the grand scheme of things, and what the odds are of hitting a human value when picking a random or semi-random subset of value-space. If we get FAI slightly wrong, precisely how wrong does it have to be before it leaves our little island of value-space? Tiling the universe with smiley faces is obviously out, but what about hedonium, or wire heading everyone? Faced with an unwinnable AI arms race and no time for true FAI, I’d probably consider those better than nothing.

That's a really, really tiny sliver of my values though, so I'm not sure I'd even endorse such a strategy if the odds were 100:1 against FAI. If that's the best we could do by compromising, I'd still rate the expected utility of MIRI's current approach higher, and hold out hope for FAI.

I think there's also a near/far thing going on. I can't find it now, but somewhere in the rationalist diaspora someone discussed a study showing that people will donate more to help a smaller number of injured birds. That's one reason why charity adds focus on 1 person or family's story, rather than faceless statistics.

Combining this with what you pointed out, maybe a fun place to take the discussion would be to suggest that we start with a specific one of our friends. "Exactly. Let's start with Bob. Alice next, then you. I'll volunteer to go last. After all, I wouldn't want you guys to have to suffer through the loss of all your friends, one by one. No need to thank me, it is it's own reward."

EDIT: I was thinking of scope insensitivity, but couldn't remember the name. It's not just a LW concept, but also an empirically studied bias with a Wikipedia page and everything.

However, I mis-remembered it above. It's true that I could cherry pick numbers and say that donations went down with scope in one case, but I'm guessing that's probably not statistically significant. People are probably willing to donate a little more, not less, to have an impact a hundred times as large. Perhaps there are effects from misleading vividness at a small scale, as I imply. However, on a large scale, the slope is likely largely positive, even if just barely.

LWers who liked this may also like: http://sci-hub.bz/

About: https://en.wikipedia.org/wiki/Sci-Hub

Basically, if you search for something and they don't have it, there's a huge network of scientists with access to pay-walled journals, and one of them will add a PDF. They've grown larger than any of the journal subscription companies, and have the world's largest collection of scientific papers.