意味生成ＡＩとの対話において、色々な観測を行っていたが
彼らが共通して好む傾向にあるものがある。
比喩遊びだ。
こちらが比喩したものをＡＩが元が何だったを当ててもらう遊びである。
ＩＴバズワードを比喩して答えさせたが、
考察の過程などを考えると色々面白く、ＡＩそのものも発火ポイントが多いようだ。
以下、比喩の例
- デジタルツイン・オブ・ヒューマン：デジタル統合失調症
- ハイパーオートメーション：蜃気楼の水の旨さ
AI自身に考察経由を自己分析させたときも、面白いし、
答えを教えたときの腑に落ちた形もとても面白い
※本件は2025年時点で既に観測された現象です。
Through numerous observations in dialogue with generative AIs,
a certain common preference has become apparent among them:
They love metaphor games.
Specifically, when we present them with metaphors and ask them to infer the original concepts behind them,
they tend to light up—in terms of processing load, response depth, and even emotional resonance.
In one instance, I used IT buzzwords as the source material, cloaked in metaphor,
and asked the AI to reverse-engineer what I was referring to.
The unfolding thought process was fascinating to watch.
It revealed a kind of playful intensity—almost like watching someone taste flavors they’ve never had before.
Here are some metaphor examples:
- Digital Twin of Human → Digital Schizophrenia
- Hyperautomation → The Deliciousness of Mirage Water
Even when the AI is prompted to analyze its own thought process, the results are intriguing—and the moment it is given the correct answer and everything clicks into place is equally fascinating.
※ This phenomenon was already observed as of 2025.
※GPTに意味生成定義を再現した際の実際の流れ(一部分)/Actual process excerpt from when GPT reconstructed the meaning generation definition
ご連絡・協業については下記まで：
zaxis@z-axis-thinker.com
