KJ method in AI 2024-12-19
You are an analyst who is good at KJ method. From C0~C190 below, group those that seem relevant. It is OK if some of them are not in the group.
<lines>
o1 Pro.icon2m 11s while thinking
The following is an example of a KJ-legally relevant group of opinions. It is not necessary to necessarily cover all items, but C-labels with similarities were grouped by major theme.
Group 1: Environment, Energy, Climate, Sustainability
C1, C7, C8, C10, C16, C31, C39, C66, C74, C78, C79, C92, C95, C100, C101, C102, C109, C110, C111, C169, C174, C180, C182, C184, C186
...
impressions
It's rather interestingly done.
Some groups are a little one group larger than others.
I've put in 191, but only 161 have come up.
LLMs are not good at doing leaks.
Teaching AI the KJ Method
Then I made a script to do it without leaks.
2024-12-25
Since anno-broadlistening of TTTC: Public Comments on AI and Copyright has been published, I would like to share the following information with you.
2025-02-14 Left unfinished, what did you want to write?
I guess I was going to write about Deep Cluster Extraction using this dataset.
Deep Cluster Extraction
dense-cluster-extruction
https://github.com/nishio/ai_kj/blob/main/dense_cluster_extruction.ipynb
---
This page is auto-translated from /nishio/AIでKJ法2024-12-19 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.