Landing abstract concepts on concrete facts
After the concept "we are the third time," which seemed to be useful in an abstract layer, was created, we tried to apply it directly to AI
Review what happened the first two times and apply it to the third time.
but gave an answer that made no sense.
I guess I'll have to delve into why I'm feeling this non-verbal sense of "streakiness" later.
In response, I let the AI list the relevant specifics.
List what historical events occurred in the first two sessions.
__BELOW_IS_AI_GENERATED__
抽象的な概念を具体的な事実に着地させる 2023-09-06 21:32 omni.icon
digest.
It discusses the importance of putting abstract concepts into concrete facts.
Relation to related fragments.
This note is related to the fragments "[Is accuracy lost when you reduce the level of abstraction?", "Digging down to specifics", "The importance of 'digging down' and 'building up' knowledge", "Digging down and circular references", "Exploration Net (fireworks) study group", " The Engineer's Art of Intellectual Production: A Reaction Summary," "Jiro Kawakita Doesn't Tell Us to Abstract," "An Experiment to Make GPT4 Supplement Context," and "nishio.icon" are all relevant. These fragments discuss the importance of putting abstract concepts into concrete facts, how to do so, and the results.
deep thinking
Putting abstract concepts into concrete facts is important to deepen understanding and to apply knowledge to real phenomena. However, in the process, information may be reduced and generality may be lost. Therefore, an appropriate balance between abstract concepts and concrete facts must be maintained. title of thought.
Balancing the Abstract and the Concrete: Deepening and Applying Knowledge."
extra info
TITLES: ["Is accuracy lost when abstraction is reduced?"], "Digging down to the concrete", "The importance of "digging down" and "building up" knowledge", "Digging down and circular references", "Exploration Net (Fireworks) Study Group", "Abstract Concepts Page", "The Engineer's Art of Intellectual Production Reaction Summary", "Jiro Kawakita is not saying to abstract.", "Experiments to make GPT4 compensate for context."]
generated: 2023-09-06 21:32
---
This page is auto-translated from /nishio/抽象的な概念を具体的な事実に着地させる 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.