Scrapbox Link
thumbnail (i.e. miniature image)
https://gyazo.com/d9e8cfd05228f6b1ddca7e8f9ffc588b
Scrapbox's concept of links is unique and much broader than what most people think of when they think of the word "link.
The expanse of it is beneficial to intellectual production, but it is hard to convey, so I will try to explain it.
https://gyazo.com/a24b19790f393891a467c000245cf82e
This is the kind of process that most people associate with the word "link."
1: Document A is present.
2: Document B is present.
3: Create a link between A and B
From the perspective of someone familiar with Scrapbox, this is a very narrow way of thinking
Just part of the link.
This assumes humans know about A and B in advance, right?"
A link that can only be established in such a situation.
The processes that people familiar with Scrapbox refer to as links include
https://gyazo.com/d9e8cfd05228f6b1ddca7e8f9ffc588b
1: Document A is present.
2: Lots of "links that do not assume that there is an existing page to link to" grow out of A.
3: Each link searches for similar links by ambiguous search
4: Similar link found!
5: By tracing it, we can find document B.
Also called "2-hop links" for the mechanism by which links are connected to each other, resulting in a connection being found between the underlying documents. The probability of this style of link occurring increases with the amount of links already in place
https://gyazo.com/89f0b5d587518206e9ba159873941070
I'm wondering if this could be realized not as an "interface that humans see and operate" but as an "API that LLMs can operate".
---
This page is auto-translated from /nishio/Scrapboxのリンク using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.