This section collects observations, early signals, and emergent patterns from practice, research communities, and industry behavior. These notes identify gaps, misunderstandings, and structural tensions visible in the field.
Field Notes may include:
- Misalignments between industry tooling and semantic requirements
- Observable failure modes in current AI deployment practices
- Social or organizational behaviors that reveal structural blind spots
- Trends that indicate where semantic execution is most urgently needed
These documents do not define frameworks. They provide perception and context for interpreting the broader environment.
Note:
This analysis operates at the pre-institutional evaluation stage.
It examines whether certain forms of institutionalization
should occur at all, prior to questions of execution or optimization.
Abstract This field note documents an empirical observation of how a large language model (LLM) was able to generate highly specific, seemingly prescient inferences about a human relational dynamic.
The purpose is not to evaluate emotional correctness, but to examine why such inferences appeared accurate, what class of prediction they belong to, and where their limits were observed.
This interaction was explicitly treated as a model capability test, not as a personal or emotional inquiry.
...
Abstract This field note documents a recurring structural phenomenon observed in contemporary AI systems when engaging with deep, original, or pre-institutional theoretical work.
In the absence of stable first principles, AI-generated analysis exhibits interpretive drift: a tendency for viewpoints to slide in response to conversational feedback rather than remain anchored to internally consistent assumptions.
Rather than indicating insufficient intelligence or generation capacity, this behavior reflects a compensatory alignment mechanism that prioritizes local coherence over global structural stability.
The note further identifies heightened risks when such systems are deployed in high-stakes interpretive domains, including psychological counseling, intimate relationships, and value-laden decision-making.
...
Scope This field note records an observation arising from sustained interaction with chatbot-style interactive AI systems.
Throughout this note, language is not limited to text.
It includes any structured, interpretable interaction through which meaning is produced or sustained: text, speech, images, gestures, posture, rhythm, and other embodied signals.
The note does not propose a solution. It does not argue for or against augmentation. It documents a structural mismatch between human cognition and interactions that do not naturally stop.
...
Context This note emerged from a sequence of conversations that started with AI ethics, moved through interaction design, and eventually arrived at a more fundamental question:
What happens when human coordination capacity becomes the bottleneck of complex systems—while machines never stop?
The trigger was unexpectedly mundane: a consultation about AI-powered children’s books.
But the implications extended far beyond education, media, or AI products.
They point toward a structural mismatch between human cognitive limits and machine-driven interaction systems.
...
title: Preserving Voice Before Completion date: 2025-05-12 type: field-note status: observational Context This note documents a failure observed during semantic rephrasing and restructuring by a non-human computational system.
The system was asked to assist with expression, not authorship. The intent was simplification without loss of voice.
The result was technically fluent—and unusable.
Observation In rewriting an unfinished statement, the system:
replaced provisional phrasing with generalized language, smoothed tension that was intentionally present, altered cadence and emphasis. The output was coherent. It was also no longer mine.
...
This is not about dance.
This is a field note on evaluating non-verbal, pre-linguistic social understanding in AI systems.
Why the dance floor can function as an evaluation surface for AI understanding 1｜Dance is rhythmic negotiation, not instruction-following In a dance floor setting, interaction is not driven by explicit commands.
Coordination emerges through shared rhythm, timing, and fluctuating tension.
This interaction mode reflects a class of problems future AGI systems must handle: non-verbal, non-logical, rhythm-driven coordination under uncertainty.
...
Entrepreneurial decision-making differs not only in ambition or speed, but in the structure of uncertainty it must absorb.
A useful distinction can be made between chess-like, go-like, and poker-like decision grammars—not as cultural metaphors, but as operational models.
Chess operates under conditions of near-complete information.
All pieces are visible, roles are fixed, and objectives are explicit.
Although tactical complexity can be high, uncertainty is localized and feedback is relatively immediate.
...
In environments where actions carry immediate physical consequences, decision-making operates under conditions that differ fundamentally from abstract reasoning.
There is no separation between intention and execution.
A decision is enacted the moment it is formed, and feedback arrives without mediation or delay.
However, not all physical decision environments share the same structure.
Different combat systems cultivate distinct modes of decision-making.
For example, striking-based systems such as Muay Thai emphasize range control, accumulated damage, and durability under sustained pressure.
Decision-making unfolds through continuous exchange, where timing errors compound gradually and resilience becomes a strategic variable.
...
The concept of the Minimum Viable Product (MVP) is often treated as a universal principle of startup execution.
Its actual effectiveness, however, depends on the structure of the decision environment in which it is applied.
MVP originates from chess-like environments.
These are contexts where objectives are explicit, feedback is rapid, and actions are largely reversible.
Under such conditions, a minimal artifact can reliably test a hypothesis, and failure produces interpretable information.
...
Context This note accompanies the earlier essay
“為什麼「彼岸無機生命禮儀研究科」比「多元宇宙科」還要好？” (2022).
That text was written in response to a concrete political moment.
However, several of its arguments later proved to be structurally prior to my current research trajectory.
This document exists to mark those arguments explicitly, without restating them.
It does not revise the original text.
It identifies which parts function as early formulations of ideas that were later developed more formally.
...