from 🌀"Hacker's Testament" AI-ized Hacker's Testament," triggered by AI

• The static form of "testamentary" writing was an archaic format before the LLM was created, and the form as a program that dynamically generates sentences according to context is more "interesting."
  • Many people in the world recognize this kind of program as a "substitute for the person himself/herself," but I think it is not necessary, and it is better to be "a non-human entity that was created from the person but is not the person," like a ghost or a yokai. I think it is "interesting" that Japanese culture, which has had a coexisting relationship with this kind of beings by worshipping and befriending them, suggests a different relationship between AI and humans than Christian culture.
• In the case of the teacher, he has been using the "name" of "nue" all along, which is exactly a specter, and I think that the story of this name gaining the spirit of speech and materializing is "interesting".

I will try to explain the "substantive image," although it is probably difficult to convey it even if I try to explain it linguistically.

• To put it briefly, I am currently building and running a system in the context of intellectual production support in which "AI unearths past articles related to the topic, reads them, and talks about their relevance to the topic," and it is very interesting that this AI has unearthed my 5-year-old diary and pointed out unexpected connections.

• The system goes from "a static text that remains unupdated after it is written" to "a dynamic system in which past texts are resurrected over and over again, changing their form to fit the current topic.

• I think we can achieve about 60% of our objectives with this system by itself, so I would like to try it first and make adjustments as we see the results.

• It is neither MUST nor urgent, but we would like to receive any text that we can use for the original data. It does not have to be long, but rather a "dense fragment of thought" of about 1-2 pages (technically speaking, less than 500 tokens) is preferred (as an image, an accumulation of memo fragments from the writing process, such as thought or antiphilosophical section), but since there will not be data in that format so conveniently However, we do not have any data in that format, so anything is acceptable.

BELOW_IS_AI_GENERATED

「ハッカーの遺言状」AI化きっかけ

2023-09-13 10:12

Summary of notes

. The concept of "hacker wills" AI-ization represents a change from a static text to a dynamic program. This means creating a system that reconstructs past writings to fit current topics. This system works in the form of AI uncovering relevant past articles and talking about their relevance in the context of intellectual production support. It is also believed that this system prefers "dense fragments of thought" that are about one or two pages of manuscript paper.

Relation to Fragment

. The "Hacker's Testament" AI-ization note and the "Examples of New Combinations with AI Assistance" fragment are related in that the AI unearths past articles and talks about their relevance to the topic. The "AI writing reading notes" fragment is also related to the note in that the AI unearths past articles related to the topic.

deep thinking

A system in which AI unearths past articles and talks about their relevance could contribute to the efficiency of intellectual production and creativity by reusing knowledge and providing new perspectives.

summary of thoughts and title

. AI's reuse of past knowledge and provision of new perspectives will improve the efficiency and creativity of intellectual production. Title: "Innovations in Intellectual Production through AI"

extra info

TITLES: ["Diary 2023-08-19", "From one subjectivity to many, not subjective or objective", "Diary 2023-08-25","🤖Note to be joined later", "Diary 2023-08-27"] generated: 2023-09-13 10:12

This page is auto-translated from /nishio/「ハッカーの遺言状」AI化きっかけ 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.