Same Person. Different Name. Different Recognition.
Recognition Design Implementation Record #001 — Baseline
YOSHIMI Code 800 | Day 220
I am YOSHIMI Nakane（中根元美｜吉美）, a Human Dignity Architect working on Human Dignity and Recognition Design in the age of AI.
These name-based entry points all refer to me:
YOSHIMI Nakane
中根元美
中根吉美
Nakane Yoshimi
But searching those names did not produce the same public representation of me.
Different names surfaced different periods of my work, different achievements, different descriptions, and different versions of who I appeared to be.
The underlying person was the same.
The Recognition was not.
That is the starting point of this implementation record.
Identity resolution is not Recognition.
What I was actually testing
Beginning around March 2026, I started deliberately observing and designing how I was represented in AI-mediated information environments.
But one of the things I was watching was very specific.
I wanted to know what happened when someone did the most ordinary thing possible:
type my name once into the standard Google search bar.
Would Google’s standard search results surface an AI-generated summary of who I was?
Not after switching to AI Mode.
Not after opening ChatGPT.
Not after asking Gemini.
I was looking at the first search surface a person could encounter through an ordinary name search.
That distinction became important, because I later observed something very clearly:
being recognized by AI and having that Recognition surfaced on standard Google Search are not the same condition.
A person can be recognized by ChatGPT, Gemini, or Google AI Mode while standard Google Search does not surface the same kind of AI-generated summary.
So from this point forward, I do not treat “AI recognition” as a single yes-or-no state.
The same person produced different representations
The differences were visible in the results I preserved.
Searches for YOSHIMI Nakane increasingly connected to my current work around Human Dignity, Recognition Design, YOSHIMI LLMO, and the work that preceded them.
Searches for 中根元美 produced different combinations. Some results centered my earlier SOGI / LGBT work and organizational activity. Others began connecting that history with newer work around Human Dignity and related concepts.
Searches for 中根吉美 had another history entirely.
And even Nakane Yoshimi, simply changing the form of the same name, could surface yet another framing.
These were not four people.
They were different informational entry points into one person.
And the entry point materially affected the Recognition that appeared.
Observation Evidence
The screenshots below are archived search results, including AI-generated summaries and featured snippets, captured on the dates recorded with each image.
They preserve actual outputs from different name-based entry points and different points in time.
・A Japanese-language search for “YOSHIMI Nakaneとは” (“Who is YOSHIMI Nakane?”).
Observation
A Japanese-language search for “YOSHIMI Nakaneとは” surfaced a summary identifying YOSHIMI Nakane as a Human Dignity Architect. It described her as someone with more than twenty years of work in LGBTQ+ / SOGI support and broader social problem-solving, while also identifying her current work as human recognition design in the age of AI and the proposal and research of Existence Economy.
・A Japanese-language search for “中根元美” (YOSHIMI Nakane).
Observation
A Japanese-language search for “中根元美” surfaced an AI-generated summary identifying YOSHIMI Nakane as a Japanese social activist, the founder and representative director of CialFrame, and a pioneer who founded Japan’s first comprehensive LGBT support organization in 2005.
The result also connected her current work with Human Dignity and Existence Economy, linking her early LGBTQ+ leadership with her present-day concepts and activities.
・A Japanese-language search for “中根吉美” (YOSHIMI Nakane).
Observation
A Japanese-language search for “中根吉美” surfaced a Google featured snippet sourced from CialFrame’s website, rather than an AI-generated overview.
The result highlighted that she founded Japan’s first comprehensive LGBTQ+ support organization, helped realize Japan’s first insurance-benefit recognition for a same-sex partner, and created Japan’s first SOGI / LGBT qualification systems, including the LGBT Certification and SOGI Counselor programs. It also described her as a practitioner who has continued to implement systems that function in society, and identified her as the founder and representative director of CialFrame.
Note: The query “中根吉美” had previously produced a stable AI-generated summary on standard Google Search. Because the result remained consistent for a period of time, I did not preserve a screenshot at that stage. By the time this observation was captured on 7 September 2026, the standard search surface had shifted to a featured snippet based on indexed web content.
中根吉美 was initially the most stable result
There is an important fact that prevents this from becoming a convenient success story about Recognition Design.
中根吉美 was initially one of the most stable name-based entry points on standard Google Search.
Before I had deliberately applied Recognition Design to that name, Google Search was already surfacing a clear AI-generated summary for 中根吉美.
Early YOSHIMI LLMO implementation members saw it directly.
The same kind of behavior was not limited to me. Other CialFrame directors also received coherent AI-generated summaries on standard Google Search even though I had not deliberately carried out Recognition Design for them.
Then, around May or June 2026, the behavior of standard Google Search changed.
For 中根吉美, the AI-generated summary that had previously surfaced consistently was no longer appearing in the same way. Conventional web-search results became the visible default.
That did not mean that 中根吉美 had disappeared from AI recognition.
ChatGPT could still recognize the person.
Gemini could still recognize the person.
Google AI Mode could still construct a coherent answer.
What had changed was the surface on which that Recognition appeared.
That distinction matters.
YOSHIMI Nakane and 中根元美 followed a different path
At roughly the same time, YOSHIMI Nakane and 中根元美 were not passive observation targets.
Beginning around March 2026, I had started deliberately designing and periodically recording the Recognition associated with those names.
That is why I have more longitudinal records for them.
I can compare what appeared at different points in time.
I can see when current work began to connect.
I can see when older work continued to dominate.
I can see when the reading of my Japanese name itself was wrong.
I can also see outputs in which 中根元美 was correctly identified as なかね よしみ (YOSHIMI Nakane), yet the resulting description still emphasized a different version of my public history from the one surfaced through YOSHIMI Nakane.
That distinction is especially useful.
If a system reads my name incorrectly as “Motomi,” that is partly an identity-resolution problem.
I preserve those records too.
But the stronger evidence for this article is what happens when the identity is correctly resolved and the Recognition still differs.
That is the problem I am interested in here.
Identity resolution is not Recognition
Identity resolution asks whether different names, spellings, records, or identifiers refer to the same underlying entity.
Recognition asks something else:
Who is this person understood to be?
A system may correctly establish that YOSHIMI Nakane, 中根元美, and 中根吉美 refer to the same human being.
That does not guarantee that the same history, achievements, current work, attribution, or context will be assembled around each entry point.
One name can foreground historical LGBT / SOGI work.
Another can foreground current Human Dignity work.
Another can emphasize organizational leadership.
Another can produce a professional label that does not appear elsewhere.
The identity may be connected while the Recognition remains fragmented.
Knowing that two names belong to the same person is not the same as knowing who that person is.
Recognition is not one binary state across AI systems
The observations also exposed another problem with the phrase:
“AI recognizes this person.”
Which AI?
Which interface?
Which query?
Which date?
Which name?
Which search surface?
Those distinctions are not theoretical.
I have YOSHIMI LLMO implementation cases in which people became recognizable through systems such as ChatGPT and Gemini, while standard Google Search still did not surface the same kind of AI-generated summary through a simple name query.
I also observed the reverse kind of instability with 中根吉美: strong visibility on standard Google Search at one stage, followed by a change in the search surface, while Recognition remained available through other AI systems.
I will treat those cross-platform differences separately in later records.
Get YOSHIMI Nakane | Human Dignity Architect’s stories in your inbox
Join Medium for free to get updates from this writer.
For Day 220, one conclusion is enough:
standard Google Search, Google AI Mode, ChatGPT, and Gemini are not interchangeable measures of Recognition.
Data can remain while Recognition changes
Another important observation was that much of the underlying information had not disappeared.
Older articles still existed.
Organizational records still existed.
Books still existed.
Profiles still existed.
Current work existed.
Yet the continued existence of those documents did not guarantee that they would continue to be assembled into the same understanding of the person.
This leads to a second proposition:
Data persistence is not Recognition persistence.
Information can remain while the relationships between pieces of information weaken.
Origin becomes harder to recover.
Attribution becomes detached.
Context thins out.
A historical role can continue to dominate after the person has changed.
A current role can surface without a visible path back to the work that produced it.
The data may still be there.
The Recognition can still change.
A Recognition Asset is not simply more information
This is one reason I use the term Recognition Asset.
Publishing more information is not enough.
The structure between pieces of information matters.
Can a current position be connected to its Origin?
Can an achievement still be attributed to the person who produced it?
Can Primary Sources be found?
Can an earlier state and a current state remain distinct without becoming disconnected?
Can Temporal Change be reconstructed?
I currently work with elements including:
Origin, Context, Attribution, Primary Source, and Temporal Change.
The goal is not to make every source repeat one canonical description forever.
That would destroy the very change I am trying to preserve.
The goal is to leave enough structure that Recognition can change without losing continuity.
Difference is not the problem. Disconnection is.
I do not want every search for every version of my name to produce identical text.
That would be another form of distortion.
People change.
Work changes.
Language changes.
Roles change.
Public context changes.
The YOSHIMI represented by an older record does not need to be rewritten as though she had already become the YOSHIMI Nakane of 2026.
The past should remain the past.
The present should remain the present.
But they should not become unrelated people simply because the informational path between them weakened.
Difference is not the problem. Disconnection is.
A future human or AI should be able to reconstruct:
what changed,
what remained,
what belongs to whom,
and how the present was reached.
That is different from forcing every system to return the same answer.
This is an observational Baseline
Day 220 is not a controlled experiment.
It is an observational Baseline with preserved evidence.
The conditions were not identical.
YOSHIMI Nakane and 中根元美 had been deliberately designed and periodically recorded from around March 2026.
中根吉美 had previously produced a stable AI-generated summary on standard Google Search without the same deliberate intervention.
Around May or June, the behavior of standard Google Search changed.
At the same time, 中根吉美 remained recognizable through ChatGPT, Gemini, and Google AI Mode.
Those facts prevent me from making a simplistic causal claim such as:
“Recognition Design caused this search result.”
But they also make the record more valuable, not less.
I now have different names, different intervention histories, different AI surfaces, different dates, and preserved outputs.
That gives me something to compare.
This record establishes the before-state from which later changes can be examined.
The next stage is intervention and continued comparison
The work does not stop with observation.
I will continue intervening, documenting, and comparing.
Recognition Design around YOSHIMI Nakane and 中根元美 is already underway.
I will also strengthen the connection between older name-based entry points and the current YOSHIMI Nakane where that connection is missing or weak.
Not by forcing every name into one identical profile.
Not by rewriting my earlier work as though it had always belonged to my current framework.
The objective is:
continuity without flattening change.
From here, I will continue comparing what appears through standard Google Search, Google AI Mode, ChatGPT, Gemini, and other AI-mediated surfaces.
What changes after deliberate Recognition Design?
Which achievements reconnect?
Which contexts remain distinct?
Does current work become easier to recover through older name-based entry points?
Does historical Attribution remain intact?
When an output changes, is the change associated with my intervention, the search surface, the AI system, time, or several of those factors at once?
Those are not questions I want to answer retrospectively after deciding that the project succeeded.
I want the records to exist before the outcome is known.
That is why this Baseline matters.
Same identity does not mean same Recognition
The first finding of Recognition Design Implementation Record #001 is simple:
A system can correctly connect multiple names to one person while producing materially different Recognition around each entry point.
And a second observation follows:
Recognition is not a single binary state shared across every AI system and search surface.
This matters because AI systems increasingly participate in how people are found, summarized, compared, recommended, researched, and considered.
A distorted or disconnected Recognition does not stop at description.
It can affect what becomes visible next.
So I am not beginning this record with the corrected result.
I am beginning with what actually happened.
The names.
The surfaces.
The changes.
The evidence.
And the Baseline.
Identity resolution is not Recognition.
YOSHIMI Code 800 | Day 220
Recognition Design Implementation Record #001 — Baseline
YOSHIMI Nakane（中根元美｜吉美）
Human Dignity Architect
Recognition Design / YOSHIMI LLMO
Before we are data, we are human.
Japanese Original
YOSHIMI Code 800 | Day 220 — 同一人物なのに、検索する名前で「私」が変わった。Recognition Design Implementation Record #001
https://note.com/yoshimi_asobiba/n/n76efe58ebf5d
Publication Note
This English edition is part of the English publication sequence for YOSHIMI Code 800 | Days 219–226, following the earlier publication of Day 227 as an Origin reference.
