Can't remember where I got this from: "The author starts with a skeleton of an argument and turns it into prose. Your job is to take the prose and excavate the skeleton."
If the excavated skeleton is messy, either a) you're a bad reader, or b) the author is a shit thinker.
Conversation
Replying to
Why can't we just trade skeletons. so many encoding x decoding happening.
Been wondering about "how can we enable lossless transfer of information between heads?"
2
3
Yeah, I'm thinking a lot about this as well. Have a lot of scattered notes about compression/decompression, narration, serialization, etc. Leaving some raw notes here, would love your ideas? reganmian.net/blog/2010/04/1
1
2
My analogy was "Two amazingly brilliant AI systems, that can only communicate with each other using morse code"... The need to take complex knowledge structure and "serialize"/"deserialize" (like modem). Wonder if Zettleconversations can be one answer? /cc
1
2
Also fascinated by all kinds of visual representations, maps, argument maps, concept maps, etc. It feels like they have a lot of potential, but I'm not sure if they're useful in transferring knowledge, as much as the process of making them. There is some interesting research...
2
1
by Collide group at Duisburg to use concept maps for collaborative learning - each participant fills out a concept map, a system automatically diffs and uses info to prompt them in collaborating (this person knows more about A than you, ask him about it).
2
There has been a ton of research on concept maps, here are just some refs (I have the PDFs of these, send me an email and I'll share). Unfortunately I can't find exactly the paper I was thinking about - seems like Sven Manske is talking about it here sci-hub.tw/10.1109/ICALT..
