Diary 2023-09-04
I don't remember writing this in the past? Extending the Red Link with AI], data from a few days ago is fine, but if you want to develop what you're thinking about now, it's not good if the data you wrote just now is not included.
5 years ago
good
Maybe once you pull out all the [Blind spot card with no picture yet
Comment below β
.
Only one π per day emerges.
β
Could something bad happen?
Let's try it.
People who, as soon as they create a certain function, immediately try to do it in an amount that is an order of magnitude higher than expected.
If you try it and it doesn't work, you'll just say, "I guess it doesn't work after all.
Great discovery if, contrary to expectations, there are no problems.
Ten times more productive than expected
Hmmm, died with an error I don't understand...
A lot of pages are processed together, so a lot of stuff is lost when things get bumped along the way.
I'm printing to the console, but not in a form that I can paste straight into Scrapbox.
I've only been able to salvage the last three because they were drifting away at the limit of the console.
That, or I should back up the JSON locally for each page processed, and then I can import it and reflect it back to where it was done.
Hmmm, it would be a shame if it falls during the 3 days I can't get back in front of my computer, but what shall I do?
If you messily surround it with TRYs and retries, you could get charged for an infinite amount of time and not be able to stop until you get home when another unexpected error occurs.
Well, if there is no error in the current trial, I would judge that it is "unlikely to cause an error in common sense use" and operate the system as it is, and if an error occurs, I would try to determine the cause.
I'll do the API call with a SLEEP in between.
If a large number of links are added at once
No, don't you want it?
I know what caused it.
[Try to pursue/abandon symmetry]
There's a bug in the API URL creation for links containing /, and it's rubbing off on the Scrapbox API, not the GPT4 API.
I have been thinking that there must be a way to combine Scrapbox with AI that utilizes the unique features of Scrapbox rather than treating it as just a bunch of text, and I think Pioneer Mode is just that. Instead of reading a bunch of text and generating text
Extend the network of knowledge by reading the network of knowledge
omni.icon
Summary.
A thought is composed of atomic parts, which are combined through "glue. Source code is not an end in itself, but a means to a higher purpose. It "cherry-picks" and adopts useful insights from different purposes and contexts. Improvements and problems with note-taking and commenting are also identified. In addition, there is a discussion of working with AI and extending the network of knowledge.
Relation to Fragment.
The note describes the atomistic components of thought and their combination,
Fragment X embodies this using the concepts of "multi-head" and "branch".
In addition, the issues of note-taking and commenting noted in the notes are discussed in Fragment X, as well as the splitting and integration of pages.
deep thinking
The relationship between the notebook and fragment X suggests that it is possible to generate diverse perspectives and ideas by combining the atomistic components of thought. In addition, the "glue" for combining these parts is an important element that supports flexibility and creativity in thinking. Furthermore, the process of combining these parts, like source code, is not an end in itself, but can be viewed as a means to achieve a higher-order goal.
summary of thoughts and title.
The atomistic parts of thinking and their combinations generate diverse perspectives and ideas and support flexibility and creativity in thinking. The process of combining these parts is a means to a higher purpose.
__BELOW_IS_AI_GENERATED__
ζ₯θ¨2023-09-04 2023-09-05 01:52 omni.icon
extra info
TITLES: ["Diary 2023-08-29", "Nodal point of thought on Scrapbox and AI 2023/8/16", "π€Scrapbox and AI 2023/8/16"]
generated: 2023-09-05 01:52
---
This page is auto-translated from /nishio/ζ₯θ¨2023-09-04 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.