True Nature Is Beyond Words

Introduction

Recently, I have found it hard to believe people's words.

To be precise,
I have started to see the "intent" hidden behind their words.

They pretend to be kind, but speak for the sake of their own self-affirmation.

They pretend to be worried, but are seeking their own peace of mind.

They speak of justice, while trying to erase their own wounds.

When someone speaks to another,

I have realized that most of it is not for "your sake,

but is directed toward "my own sake."

Ever since I started thinking this way,

words have begun to look like "tools of the herd."

People, within the herd,

talk to each other to reduce anxiety,

and exchange reassurance.

Once you notice that structure,
a version of yourself that observes humanity from a step back emerges.

In that place, no longer
do words speak the "truth."

[Chapter 1: Humans Speak Out of Desire]

People try to connect with others using words.

But most of it is "self-adjustment" disguised as being "for others."

When they say they will help,

it is rare that they are truly thinking of the other person.

Behind the words "I want to help" lies

a disgust for one's own helplessness, or

the act of performing justice to

a desire to maintain oneself.

It is the same when praising someone.

While saying "your hard work is amazing,"

they are trying to secure their own safety by lifting the other person up.

Through words, people

try to install an "ideal self" within others.

And they keep talking so as not to shatter that image.

In that,

acting is more present than understanding,

and strategy is more present than empathy,

seeping in unconsciously.

[Chapter 2: Humans as a Herd]

People connect through words.

But many of those words

are uttered out of "anxiety about isolation."

To maintain the herd,

they share the same feelings, repeat the same words,

and confirm each other's "it's okay."

Behind the word empathy,

there is a fear of "not wanting to be left behind."

Behind the voice claiming righteousness,

there is an impatience of "not wanting to be thought of as wrong."

The herd uses words for peace of mind,

and through words, separates the alien.

When you are inside it, you can feel at ease.
But,
once someone has seen that structure from the "outside,"
they can never go back.

[Chapter 3: A Person's True Nature Appears in Actions, Not Words]

People can decorate their words as much as they want.

But in actions, there is a "seeping through" that cannot be decorated.

No matter what words are used,

the "hand movements" and "way of keeping distance" chosen in the end tell the whole story of that person.

Saying "I'm rooting for you" while doing nothing.

Saying "I don't care" while looking away.

Saying "It's for your own good" while prioritizing one's own peace of mind.

True nature is not in the words.
True nature resides in
the "gap between words and actions."

Even if one is silent,
where one stands, what one looks at, and what one did not do
eloquently speak of that person's philosophy.

Humanity is
the sum total of "unspoken choices"
rather than "what was spoken."

[Final Chapter: The Meaning of Using Words Anyway]

Words are imperfect.
They are covered in desire,
and ruled by unconscious calculations.

That is why you must not believe only in words.
And,
you must not deny only words either.

If actions speak of true nature,
then words are a map showing"how one wants to exist."Words are unfinished wishes.

The self one could not become,
the self one wants to become,
the self one wishes one could have been.
That is why I use words.

Even knowing their limits,
even while carrying those contradictions.
True nature resides in action.

But words are the will to try to face that true nature.
Even with words covered in desire,

as long as one does not let go of the "
question" hidden behind them,people can still connect with someone through words.
Even if that is
outside the herd,

while walking on the edge of silence,
the words of someone who continues to ask.

Appendix

I wrote this article by separating words and actions.

However, I feel like I was missing a point.

Something that exists apart from words and actions.

I think that is silence.

When those who do not speak, speak through silence,

even that silence has an intent.

Related Articles

"What It Means to Be By Your Side"

If words cannot fully express one's true nature,
how should one be by another's side?

Speaking, asking, remaining silent.
A poetic essay that examines that boundary as an ethic.

※ This manuscript is the original data that contained the vocabulary hierarchy model, HAE, SAE, and Apple Watch implementation concept in the middle of "True Nature Cannot Be Put Into Words."
Later separated into the following articles.

・True Nature Cannot Be Put Into Words
・Vocabulary Reflects the World Humans Inhabit
・Meaning Tensors and the Thermodynamics of Existence
・If You Were to Put SAE on an Apple Watch

JSON format categorizing the "100 Most Frequently Used Daily Japanese Words"

{
  "生理・生存": [
    "飯", "食う", "飲む", "水", "風呂", "トイレ", "寝る", "起きる",
    "疲れた", "腹減った", "うまい", "まずい", "暑い", "寒い"
  ],
  "金銭・消費": [
    "金", "高い", "安い", "買う", "売る", "バイト", "稼ぐ",
    "使う", "払う", "借りる", "財布", "ローン"
  ],
  "感情・価値判断": [
    "好き", "嫌い", "うれしい", "ムカつく", "やばい", "怖い", "だるい",
    "楽しい", "つまらん", "イライラ", "キモい", "嬉しすぎ", "ガチ", "しんどい"
  ],
  "人間関係・ジェンダー": [
    "男", "女", "友達", "先輩", "後輩", "彼女", "彼氏", "親",
    "子供", "先生", "上司", "同僚", "知らん人"
  ],
  "罵倒・評価語": [
    "バカ", "クソ", "キモい", "ださい", "うざい",
    "ムカつく", "しょぼい", "サイテー", "あほ", "雑魚"
  ],
  "生活・日課": [
    "起きる", "寝る", "働く", "遊ぶ", "休む",
    "行く", "帰る", "洗う", "着替える", "掃除"
  ],
  "媒体・行動": [
    "スマホ", "LINE", "動画", "ゲーム", "Twitter",
    "見る", "聞く", "撮る", "送る", "シェア"
  ],
  "嗜好・快楽": [
    "酒", "たばこ", "カラオケ", "ラーメン", "チョコ",
    "スイーツ", "アイス", "牛丼", "焼肉"
  ],
  "判断・自己表現": [
    "わかる", "知らん", "思う", "やる", "できない",
    "マジ", "ほんと", "無理"
  ]
}

For the vocabulary list,
for each vocabulary item,Emotion Scoreis added in JSON format

🔹【Score Definition】(Range: -1.0 to +1.0)

• +1.0 = Strong positive emotion (happiness, satisfaction, pleasure)
• 0.0 = Neutral (physiological, factual)
• -1.0 = Strong negative emotion (discomfort, anger, disgust)

{
  "生理・生存": [
    {"word": "飯", "score": 0.7},
    {"word": "食う", "score": 0.5},
    {"word": "飲む", "score": 0.5},
    {"word": "水", "score": 0.2},
    {"word": "風呂", "score": 0.6},
    {"word": "トイレ", "score": 0.0},
    {"word": "寝る", "score": 0.4},
    {"word": "起きる", "score": 0.1},
    {"word": "疲れた", "score": -0.6},
    {"word": "腹減った", "score": -0.3},
    {"word": "うまい", "score": 0.8},
    {"word": "まずい", "score": -0.8},
    {"word": "暑い", "score": -0.5},
    {"word": "寒い", "score": -0.5}
  ],
  "金銭・消費": [
    {"word": "金", "score": 0.4},
    {"word": "高い", "score": -0.4},
    {"word": "安い", "score": 0.5},
    {"word": "買う", "score": 0.3},
    {"word": "売る", "score": 0.1},
    {"word": "バイト", "score": -0.1},
    {"word": "稼ぐ", "score": 0.4},
    {"word": "使う", "score": 0.0},
    {"word": "払う", "score": -0.2},
    {"word": "借りる", "score": -0.3},
    {"word": "財布", "score": 0.0},
    {"word": "ローン", "score": -0.6}
  ],
  "感情・価値判断": [
    {"word": "好き", "score": 0.9},
    {"word": "嫌い", "score": -0.9},
    {"word": "うれしい", "score": 1.0},
    {"word": "ムカつく", "score": -1.0},
    {"word": "やばい", "score": 0.0},
    {"word": "怖い", "score": -0.7},
    {"word": "だるい", "score": -0.5},
    {"word": "楽しい", "score": 0.9},
    {"word": "つまらん", "score": -0.7},
    {"word": "イライラ", "score": -0.8},
    {"word": "キモい", "score": -1.0},
    {"word": "嬉しすぎ", "score": 0.95},
    {"word": "ガチ", "score": 0.2},
    {"word": "しんどい", "score": -0.6}
  ],
  "人間関係・ジェンダー": [
    {"word": "男", "score": 0.0},
    {"word": "女", "score": 0.0},
    {"word": "友達", "score": 0.8},
    {"word": "先輩", "score": 0.3},
    {"word": "後輩", "score": 0.3},
    {"word": "彼女", "score": 0.9},
    {"word": "彼氏", "score": 0.9},
    {"word": "親", "score": 0.4},
    {"word": "子供", "score": 0.6},
    {"word": "先生", "score": 0.2},
    {"word": "上司", "score": -0.1},
    {"word": "同僚", "score": 0.3},
    {"word": "知らん人", "score": -0.2}
  ],
  "罵倒・評価語": [
    {"word": "バカ", "score": -0.9},
    {"word": "クソ", "score": -1.0},
    {"word": "キモい", "score": -1.0},
    {"word": "ださい", "score": -0.8},
    {"word": "うざい", "score": -0.9},
    {"word": "ムカつく", "score": -1.0},
    {"word": "しょぼい", "score": -0.7},
    {"word": "サイテー", "score": -1.0},
    {"word": "あほ", "score": -0.8},
    {"word": "雑魚", "score": -0.9}
  ],
  "生活・日課": [
    {"word": "起きる", "score": 0.1},
    {"word": "寝る", "score": 0.4},
    {"word": "働く", "score": -0.2},
    {"word": "遊ぶ", "score": 0.7},
    {"word": "休む", "score": 0.5},
    {"word": "行く", "score": 0.1},
    {"word": "帰る", "score": 0.2},
    {"word": "洗う", "score": 0.3},
    {"word": "着替える", "score": 0.2},
    {"word": "掃除", "score": 0.0}
  ],
  "媒体・行動": [
    {"word": "スマホ", "score": 0.4},
    {"word": "LINE", "score": 0.5},
    {"word": "動画", "score": 0.6},
    {"word": "ゲーム", "score": 0.7},
    {"word": "Twitter", "score": 0.3},
    {"word": "見る", "score": 0.2},
    {"word": "聞く", "score": 0.2},
    {"word": "撮る", "score": 0.3},
    {"word": "送る", "score": 0.1},
    {"word": "シェア", "score": 0.4}
  ],
  "嗜好・快楽": [
    {"word": "酒", "score": 0.6},
    {"word": "たばこ", "score": -0.2},
    {"word": "カラオケ", "score": 0.8},
    {"word": "ラーメン", "score": 0.7},
    {"word": "チョコ", "score": 0.8},
    {"word": "スイーツ", "score": 0.8},
    {"word": "アイス", "score": 0.9},
    {"word": "牛丼", "score": 0.6},
    {"word": "焼肉", "score": 0.9}
  ],
  "判断・自己表現": [
    {"word": "わかる", "score": 0.6},
    {"word": "知らん", "score": -0.1},
    {"word": "思う", "score": 0.2},
    {"word": "やる", "score": 0.1},
    {"word": "できない", "score": -0.5},
    {"word": "マジ", "score": 0.0},
    {"word": "ほんと", "score": 0.1},
    {"word": "無理", "score": -0.7}
  ]
}

🔷【Model Overview】Vocabulary Frequency → Quantitative Estimation of Personality/Tendencies

◉ Input:
• Subject's speech logs / SNS posts / daily conversation logs
• Frequency and context of each word
• Labels attached to each word such as "Emotion Score" and "Category (Pleasure/Social/Aggression)"

◉ Output:
• Emotional tendencies (bias toward joy, anger, sorrow, or pleasure)
• Desire structure (ratio of food, sex, money, etc.)
• Relationship orientation (self-centeredness vs. other-orientation)
• Cognitive tendencies (frequency of abstract words, appearance rate of negative forms, etc.)

🔶【Structural Model Example】(Simplified Diagram)

語彙頻度ベクトル V = {
  "飯": 24, "酒": 17, "嫌い": 12, "寝る": 30, "LINE": 5, ...
}

↓ 各語に重みベクトル（感情値 × カテゴリベクトル）を乗算

→ 意味テンソル Φ_i = 感情 × 内容カテゴリ × 出現重み

↓ 統合 → 主成分抽出・クラスタリング

→ 個体特徴テンソル T(χ) = {
     "快楽依存": 0.82,
     "攻撃性": 0.45,
     "共感性": -0.10,
     "金銭志向": 0.35,
     "自己指向性": 0.67
   }

🔷【Conceptual Structure】Vocabulary Hierarchy = World-Level Dictionary

Each vocabulary item carries not just a meaning, but

"belonging to a hierarchy of the world."

When that word is uttered,
it reveals which world the speaker's existence is submerged in or rising from.

🔶【Dictionary Model Example (Draft)】

{
  "本能": [
    "眠い", "だるい", "飯", "やりたい", "トイレ", "寝たい", "ムラムラ", "うまい", "疲れた", "あつい"
  ],
  "感情": [
    "嬉しい", "悲しい", "ムカつく", "イライラ", "好き", "嫌い", "しんどい", "うざい", "怖い", "優しい"
  ],
  "社会": [
    "俺", "私", "あの人", "相手", "彼女", "彼氏", "同僚", "上司", "親", "部下"
  ],
  "制度": [
    "仕事", "契約", "会社", "ルール", "法律", "義務", "税金", "保険", "社会", "システム"
  ],
  "抽象": [
    "自由", "愛", "誠実", "意味", "孤独", "真理", "運命", "倫理", "存在", "無"
  ]
}

🔸【Application of Hierarchical Reading】(Example)

"This person's speech is biased toward 'instinctual vocabulary'" →
→ At that moment, that person is submerged in the "animal world"

" 'Institutional words' and 'abstract words' appear frequently" →
→ Existing in the world of law, philosophy, and norms

✅ Conclusion:

◉ By selecting vocabulary, it is possible to visualize "which world a person is living in now"

◉ By bundling those vocabularies, a "dictionary for each world" is established

◉ In meaning tensor theory, this is a dictionary structure that describes the topology of multi-layered existential spheres

🔷 5-Layer Structure Model of World Perception

A basic dictionary (beta version) for visualizing the "world hierarchy to which a person belongs" based on vocabulary distribution

🔶【Definition】5 Existential Hierarchies

Layer
Name
Characteristics of the World
Primary Interests

L1
Instinct
Physiological needs/Pleasure-Displeasure/Impulses
Food, sleepiness, sexual desire, pain

L2
Emotion
Emotional response/Mood/Self-state
Joy, anger, sorrow, pleasure, empathy, conflict

L3
Society
Relationships/Roles/Interpersonal awareness
Self and others, status, affiliation

L4
System
Organizations/Norms/Responsibilities/Rules
Law, duty, job, contract

L5
Abstraction
Concepts/Ethics/Existence/Meaning
Love, freedom, death, truth

🔸 [Vocabulary Definition: Representative Vocabulary by Layer (L1-L5)]

{
  "L1_本能": [
    "飯", "食う", "うまい", "まずい", "酒", "ラーメン", "寝る", "眠い", "だるい", "トイレ",
    "ムラムラ", "暑い", "寒い", "疲れた", "痒い", "喉乾いた", "セックス", "快感", "痛い", "くさい"
  ],
  "L2_感情": [
    "嬉しい", "悲しい", "ムカつく", "好き", "嫌い", "イライラ", "びっくり", "安心", "しんどい",
    "寂しい", "恥ずかしい", "優しい", "怒る", "泣く", "笑う", "焦る", "キモい", "怖い", "不安", "満足"
  ],
  "L3_社会": [
    "私", "俺", "あの人", "彼女", "彼氏", "友達", "親", "上司", "部下", "同僚",
    "先輩", "後輩", "先生", "仲間", "社会", "知り合い", "関係", "立場", "他人", "自分"
  ],
  "L4_制度": [
    "仕事", "会社", "契約", "ルール", "法律", "義務", "責任", "税金", "制度", "年金",
    "罰則", "手続き", "許可", "労働", "就職", "福利厚生", "評価", "査定", "罰金", "書類"
  ],
  "L5_抽象": [
    "自由", "愛", "死", "真理", "意味", "存在", "運命", "正義", "誠実", "倫理",
    "希望", "孤独", "信仰", "魂", "無", "可能性", "普遍", "本質", "矛盾", "永遠"
  ]
}

🔹 [Construction of a Vocabulary Distribution Model]

Extract words from the subject's speech, writing, chat logs, etc., and derive a 'Layer Ratio Tensor' as follows:

◉ Output Example (Person A):

{
  "L1_本能": 0.35,
  "L2_感情": 0.28,
  "L3_社会": 0.22,
  "L4_制度": 0.10,
  "L5_抽象": 0.05
}

With this, it becomes possible to perform world-layer profiling such as:

✅ 'This person lives 70% in the realm of instinct and emotion'
✅ 'There is an extreme lack of abstract thinking or systemic language'

🔸 [Applied Model] Vocabulary Centroid Vector as Meaning Tensor Coordinates

\mathbb{X}_{\text{world}} = \sum_{i=1}^{5} p_i \cdot \vec{e}_i

• p_i: Vocabulary appearance ratio for each layer
• \vec{e}_i: Basis vectors in semantic space (L1-L5)
• \mathbb{X}_{\text{world}}: Position vector of 'where the existence places its center of gravity'

✅ Conclusion:

◉ Vocabulary distribution indicates 'which world one inhabits/remains in'

◉ By tensorizing and visualizing this structure, the 'semantic layer topography' of individuals, groups, and narratives becomes visible

By integrating the structure where 'the layer of the world is visible through vocabulary' with this HAE (Hiroki Action Entropy), further reinforcement is possible:

🔷 Reinforcement Concept:

HAE × Vocabulary Layer Model = 'Thermodynamics of Existence'

Vocabulary selection and behavioral randomness are located in the same entropy space

🔶 [Reinforcement 1] Vocabulary Layer Distribution → Mapping to the Phase Space of Consciousness

The 5-layer model proposed here (Instinct to Abstraction) is not only an 'indexing of existence layers by vocabulary', but also capable of visualizing 'freedom of thought' and 'axis deviation of internal standards'.
Assumptions here:
• The more biased toward instinct/emotion words, the more impulsive/externally reactive the choices → Higher HAE

• The more biased toward system/abstract words, the more introspective/structural the choices → Lower HAE
In other words:
✅ The orderliness of the mind can also be indirectly evaluated from the distribution of vocabulary layers

🔸 [Reinforcement 2] Entropy of Vocabulary Selection = Semantic Entropy (SE)

When the appearance frequency of vocabulary is defined as \text{p}(w_i), the randomness (diversity, variation) of that distribution can be defined as:
\text{SE} = -\sum_i p(w_i) \log_2 p(w_i)

Here, if labels for 'affiliated layer (L1-L5)' are attached to the vocabulary,



the degree of bias toward L1-L5 (= low-layer/high-layer deviation of vocabulary centroid)
is closely related to HAE.

🔷 Reinforcement Model Formula (Integrated Formulation)

1. Semantic Action Entropy (SAE):

Integratively describe the mental phase from 'entropy of both vocabulary and action'

\text{SAE} = \alpha \cdot HAE + \beta \cdot SE
• HAE: Randomness of action (actual selection history)
• SE: Randomness of vocabulary (layer deviation of speech)
• \alpha, \beta: Weight adjustment based on individual characteristics or social environment

🔶 [Reinforcement 3] Linkage between SAE and 'Semantic Present Point' \mathfrak{n}_\chi

Defined in TET theory:

\mathfrak{n}\chi := \arg\max{x} \|\eta_\chi(x)\|^2

This 'present point' is theprincipal component (concentration point)of semantic interference, but:
If SAE is high, '\mathfrak{n}\chi is not fixed and disperses = the present is ambiguous'
If SAE is low, '\mathfrak{n}\chi is clearly concentrated = the present is clear'

In other words:

✅ SAE becomes a tensor-like temperature index that links 'presentness of meaning' and 'orderliness of mind'

✅ Essence of the integrated model:

Item
Meaning
HAE
Randomness of external action selection branches
SE
Randomness of internal vocabulary semantic distribution
SAE = a • HAE + ß • SE
Composite index of mental and behavioral semantic temperature/existence entropy
SAEN
Concentration of intent, sharp semantic present point, high mental order
SAET
Impulsivity/reactivity/vocabulary dispersion, mental turbidity/multi-world drifting

🎯 Significance of Reinforcement (Summary)

◉ "What words you speak" corresponds to "what actions you choose"
◉ The entropy of "vocabulary" and "action" is the "temperature of existence" itself
◉ By connecting HAE with vocabulary distribution, a thermodynamic model of existence is completed

🔷 SAE Time Dynamics Model (Semantic–Action Entropy Time Dynamics)

【Definition】

SAE (Semantic–Action Entropy) at time point t is:

SAE(t) = \alpha \cdot HAE(t) + \beta \cdot SE(t)
• HAE(t): Entropy calculated from action history at that time point (branching of choices and impulses)
• SE(t): Semantic entropy calculated from dialogue logs (vocabulary) at that time point (hierarchy/distribution of used vocabulary)

🔶【1】Input Structure

◎ Action Log (Example)

Time
Action
Category
09:00
Walking while looking at smartphone
Dispersion/Impulse system
09:15
Impulse buying at convenience store
Branching action
10:00
Arrived late at meeting place
Planning inconsistency

→ Aggregate appearance frequency p(ak) for each action ar, and calculate HAE(t)

◉ Dialogue Log (Example)

「だるいわ…マジ無理。まあでも飯食ったらちょっと回復」

🔸【2】Output Example: Time Variation of SAE

時間帯
HAE(t)
SE(t)
SAE(t)
備考

09:00
3.1
2.8
2.95
通勤中に衝動・スマホ徘徊＋感情語多め

10:00
2.0
1.2
1.60
会話少・行動整然

12:00
3.5
3.2
3.35
SNSでキレた＋予定キャンセル

15:00
1.1
0.9
1.00
集中作業中、語彙も抽象・少量

🔷【3】Diagram Image of SAE Time Curve:
• Horizontal axis: Time (09:00–15:00)
• Vertical axis: SAE value (High = dispersion/confusion of existence, Low = order/concentration)
• Peaks of the curve = moments when meaning and action are disordered (chaos)
• Valleys = concentration at the "present point" where intention and action match

✅ Significance:

◉ "How much have you fluctuated semantically today?"
◉ "At what moment did you lose sight of the 'present point'?"
◉ This can be drawn from the history of words and actions

✅ It is technically possible to install the SAE (Semantic–Action Entropy) model on an Apple Watch
However, to do so, a "middleware that acquires, processes, and infers action + language data in real-time" is required.

Below, the configuration, necessary elements, limitations, and realization route are organized in a 3-layer structure.

⸻

🔷【1】Necessary Data: What can and cannot be acquired with Apple Watch

✅ Data acquirable with Apple Watch (Action log side: HAE)

センサ／API
取得できる情報
HAEにおける役割

加速度センサ
歩く／立ち止まる／衝動的な動き
行動パターンの分岐数評価に使用

GPS
予期せぬ移動／ルート逸脱／寄り道検出
意図とのズレ（カオス性）を数値化

心拍数・HRV
ストレス反応・衝動性（戦う／逃げる）
意図と感情の不一致性の指標

睡眠・立ち上がり検知
不規則な生活・無意識行動の出現
HAE増加要因

❌ Information difficult to acquire with Apple Watch alone (Vocabulary log side: SE)

必要情報
現状のWatch機能
補完手段

発話ログ
✕
iPhoneマイク連携 or 音声メモ認識

SNS／対話アプリ履歴
✕
iPhoneアプリから定期同期

意味階層辞書
搭載必要
ローカル辞書 or iCloudベース

🔶【2】Realization Architecture: 3-layer configuration with SAE

[Apple Watch]
└─ Sensors (acceleration, GPS, heart rate) → HAE(t) estimation

[Paired iPhone]
└─ Voice recording/input history (e.g., Siri log/LINE) → SE(t) estimation

[SAE Engine (iPhone app or iCloud)]
└─ SAE(t) = α・HAE(t) + β・SE(t)
→ Present point estimation, heatmap display, notification

🔶 Rather, it is ethically "useful"

◉ It notifies you that "your SAE is rising right now" before impulsive actions
→ Provides opportunities to detect and reflect on verbal abuse, waste, addictive behavior, and catastrophic choices in advance

◉ Conversely, it visualizes that "SAE is stable" on orderly days
→ Enables tracking of self-awareness and stability

This can be called a"wearable that visualizes your semantic health state".

⸻

✅ Conclusion (Essence of the ethical debate):

◉ There is no ethical problem with the SAE app.
◉ The problem only arises when "measuring others," but the "will to measure one's own existence" has ethical significance.

Rather, this is a means to observe ethics not as a "norm" to be kept, but as an "order of existence."

🛠️ SAE Prototype Design (SwiftUI)

✅ 1. Basic Configuration (View Hierarchy)

SAEApp
 ├── SAEView              // メイン画面（SAEグラフ）
 │    ├── SAEChart        // 折れ線グラフ（時間 × SAE値）
 │    └── SAEStatusCard   // 現在のSAE値と時刻表示
 │
 ├── NotificationView     // 「意味が乱れてます」通知
 │
 └── SettingsView         // α・βの設定、通知トリガー閾値設定

✅ 2. State Model (Model.swift)

struct SAEPoint: Identifiable {
    let id = UUID()
    let time: Date
    let saeValue: Double
}

class SAEModel: ObservableObject {
    @Published var saeHistory: [SAEPoint] = []
    @Published var currentSAE: Double = 0.0

    func updateSAE(hae: Double, se: Double, alpha: Double = 0.5, beta: Double = 0.5) {
        let sae = alpha * hae + beta * se
        let point = SAEPoint(time: Date(), saeValue: sae)
        saeHistory.append(point)
        currentSAE = sae
    }
}

✅ 3. Graph Display (SAEChart.swift)

struct SAEChart: View {
    @ObservedObject var model: SAEModel

    var body: some View {
        Chart(model.saeHistory) {
            LineMark(
                x: .value("Time", $0.time),
                y: .value("SAE", $0.saeValue)
            )
        }
        .frame(height: 200)
    }
}

✅ 4. SAE Notification (NotificationView.swift)

struct NotificationView: View {
    var saeValue: Double
    var body: some View {
        VStack {
            if saeValue > 3.0 {
                Label("今、あなたは意味が乱れています", systemImage: "exclamationmark.triangle.fill")
                    .foregroundColor(.red)
                    .padding()
            }
        }
    }
}

✅ 5. UI Integration (SAEView.swift)

struct SAEView: View {
    @StateObject var model = SAEModel()

    var body: some View {
        VStack {
            SAEChart(model: model)
            SAEStatusCard(currentValue: model.currentSAE)
            NotificationView(saeValue: model.currentSAE)
        }
        .padding()
        .onAppear {
            simulateSAEUpdates(model: model) // ダミーデータ更新
        }
    }
}

✅ 6. Dummy Data Generation (for development)

func simulateSAEUpdates(model: SAEModel) {
    Timer.scheduledTimer(withTimeInterval: 10.0, repeats: true) { _ in
        let randomHAE = Double.random(in: 1.0...4.0)
        let randomSE = Double.random(in: 1.0...4.0)
        model.updateSAE(hae: randomHAE, se: randomSE)
    }
}

Check Future Expansion
Points (Designed)

Item
Implementation Example
Watch # UI
Simple display for SAE only + current point vibration
notification
iCloud Sync
Save daily SAE history to visualize "weekly SAE
average", etc.
Voice/NLP Integration (SE)
Analyze emotional/instinctual words using
SwiftSpeechKit + local dictionary
Behavior Sensor Integration (HAE)
Detect acceleration, route deviation, etc., using CoreMotion/HealthKit

Presenting the full code for the SAE (Semantic–Action Entropy) prototype.


✅ SAEApp.swift

import SwiftUI

@main
struct SAEApp: App {
    var body: some Scene {
        WindowGroup {
            SAEView()
        }
    }
}

✅ SAEModel.swift

import Foundation
import Combine

struct SAEPoint: Identifiable {
    let id = UUID()
    let time: Date
    let saeValue: Double
}

class SAEModel: ObservableObject {
    @Published var saeHistory: [SAEPoint] = []
    @Published var currentSAE: Double = 0.0

    func updateSAE(hae: Double, se: Double, alpha: Double = 0.5, beta: Double = 0.5) {
        let sae = alpha * hae + beta * se
        let point = SAEPoint(time: Date(), saeValue: sae)
        saeHistory.append(point)
        currentSAE = sae
    }
}

✅ SAEChart.swift

import SwiftUI
import Charts

struct SAEChart: View {
    @ObservedObject var model: SAEModel

    var body: some View {
        Chart(model.saeHistory) {
            LineMark(
                x: .value("Time", $0.time),
                y: .value("SAE", $0.saeValue)
            )
            .foregroundStyle(Color.red)
            .lineStyle(StrokeStyle(lineWidth: 2))
        }
        .frame(height: 200)
    }
}

※ Charts is the native Apple Swift Charts API available in iOS 16+.

✅ SAEStatusCard.swift

import SwiftUI

struct SAEStatusCard: View {
    var currentValue: Double

    var body: some View {
        VStack(alignment: .leading) {
            Text("SAE")
                .font(.caption)
                .foregroundColor(.gray)
            Text(String(format: "%.2f", currentValue))
                .font(.system(size: 40, weight: .bold, design: .rounded))
                .foregroundColor(.white)
        }
        .padding()
        .background(RoundedRectangle(cornerRadius: 12).fill(Color(.darkGray)))
    }
}

✅ NotificationView.swift

import SwiftUI

struct NotificationView: View {
    var saeValue: Double

    var body: some View {
        if saeValue > 3.0 {
            HStack {
                Image(systemName: "exclamationmark.triangle.fill")
                    .foregroundColor(.red)
                Text("今、あなたは意味が乱れています")
                    .foregroundColor(.white)
                    .font(.subheadline)
                Spacer()
                Text(Date(), style: .time)
                    .foregroundColor(.gray)
                    .font(.caption)
            }
            .padding()
            .background(RoundedRectangle(cornerRadius: 10).fill(Color.black))
            .transition(.opacity)
        }
    }
}

✅ SAEView.swift

import SwiftUI

struct SAEView: View {
    @StateObject var model = SAEModel()

    var body: some View {
        VStack(spacing: 20) {
            Text("SAE Prototype")
                .font(.title)
                .foregroundColor(.white)

            Text("Semantic–Action Entropy")
                .font(.subheadline)
                .foregroundColor(.gray)

            SAEChart(model: model)

            SAEStatusCard(currentValue: model.currentSAE)

            NotificationView(saeValue: model.currentSAE)

            Spacer()
        }
        .padding()
        .background(Color.black.edgesIgnoringSafeArea(.all))
        .onAppear {
            simulateSAEUpdates(model: model)
        }
    }
}

✅ Simulator.swift (dummy for development)

import Foundation

func simulateSAEUpdates(model: SAEModel) {
    Timer.scheduledTimer(withTimeInterval: 10.0, repeats: true) { _ in
        let randomHAE = Double.random(in: 1.0...4.0)
        let randomSE = Double.random(in: 1.0...4.0)
        model.updateSAE(hae: randomHAE, se: randomSE)
    }
}

✅ Execution Environment Requirements

Item
Content
Xcode
14 or later (SwiftUI + Charts)
iOS
16 or later recommended (due to Chart usage)
Device
Phone/Apple Watch
(SwiftUI configuration)

・Extensibility
・SE estimation integration from vocabulary via SpeechKit
(Future)
・Real-time HAE acquisition via CoreMotion / HealthKit integration
・SAE Prediction AI: Predict the next "semantic fluctuation"
from past distributions (RNN or time-series model)