The current review culture has already gone beyond mere “sharing impressions” and is exposing aspects of a kind of “social pathology”.

What I’ve long felt after years of publishing on Kindle

I’ve been publishing on the Amazon Kindle Store for many years.
And through that, there’s something I’ve felt very strongly all along.

That is,many online reviews, especially the reviews attached to Kindle books, are not functioning as “reviews” in the true senseThat’s what I mean.

Of course, I’m not saying that negative reviews themselves are bad.
Rather,if a negative review is factually verifiable, I actually welcome it.
Because it can be improved.

For example,

• which statement on which page has a problem
• which explanation was hard to understand
• which claim contains a leap in logic
• which part did not match expectations

If it is written concretely in that way, the author can check it, and readers can verify it too.
That is extremely healthy as a review, and it has value.

But in reality, reviews like that are astonishingly rare.

Most are,unverifiable。
or,little more than insults。
More bluntly,the reader’s own lack of understanding, mood, assumptions, or misreading is written as if it were a defect in the work itselfProjection and the like, too.

And the troublesome thing is that they often appear **as if they were “the truth”**.
And that can influence some people who cannot see through the lie, leading them to accept the lie as fact.

Having published on Kindle for many years, I’ve felt a strong sense of unease about this point for a long time.
A review system should originally exist to create a healthy connection between the work and the reader.
But in reality,it often becomes a place that actually lowers the value of a work more than it deserves.I see this not as a mere personal complaint, but as

a kind of social pathology.The problem is not “bad reviews”

The problem is that “unverifiable reviews” circulate as if they were facts

Please don’t misunderstand this point.

I’m not saying this because I hate bad reviews.

I’m not saying it because I don’t want criticism.
What I truly think is problematic is,

the content of the criticismbeingunverifiable.

“It was a confusing book”
“The content was thin”
“It’s untrustworthy”
“What the author says doesn’t make sense”
“It’s not worth reading”

Even if comments like these are present, if they do not statewhat exactly was read and led to that judgmentneither the author nor any third party can verify them. I think it is a kind of social pathology that many people casually spew out things like this as if it were normal.

Strictly speaking, that is not a review.
It is nothing more thanthe injection of unverifiable impressions

And on the internet, this kind of thing tends to have a particularly strong effect.
Because people are more easily pulled by short, strong negative impressions than by careful examination.

Careful criticism takes time.
But sloppy denial creates an atmosphere in an instant.

I dislike this “atmosphere.” I feel it is extremely immature.
Even when there has actually been no fact-checking at all, a vague mood can form that says, “this book seems dangerous” or “this author seems suspicious.”

In other words, what is happening in many review sections now is not evaluation of the work, butcontagion of impressions.
It is the spread of a false atmosphere. People who cannot even use Japanese properly write reviews without hesitation while criticizing the book’s writing. It’s like projection.

In review sections, not only the work’s issues but also “the reader’s issues” get mixed in

Even more serious is that review sections often containthings other than the work itself.

For example,

• the reader was unfamiliar with the topic
• it differed from what they expected
• it didn’t match their state of mind at the time
• they were not prepared to read it
• their method of comprehension was off
• they had a negative bias from the beginning

In reality, situations like these naturally exist.
But they do not necessarily mean the work is defective.

Even so, in review sections,
“I couldn’t accept it” can get swapped for “this work has no value”That is a major distortion.

Of course, impressions should be free.

“It didn’t suit me” is a perfectly valid impression.
But impressions and work evaluations are not, in principle, the same thing.
Yet in today’s review sections, that boundary is far too blurry.

As a result,
the other person’s problem is displayed as the work’s problemThis is not only unfair to the author.。

It is also disadvantageous for readers.
Because what is needed as material for purchasing decisions is verifiable information.
In other words, it becomes acceptable for those in power to tell those without power, “don’t read that book because it’s a bad book.” I believe this is a kind of social pathology.

Even more troublesome is that the possibility of staged reviews, reverse stealth marketing, or malicious intent can never be completely ruled out

The reason online review culture becomes unhealthy is not just misunderstanding or misreading.

Rumors of fake reviews and unnatural review manipulation have never stopped.

To lower someone’s value, or to bring down a competitor, a person may pay someone to write an unnatural review.
Who can say that such things do not exist at all?
Of course, it is difficult for outsiders to determine whether each individual review is truly like that.

But that’s not the core issue.
What matters is,

that today’s review sections make it almost impossible to distinguish between innocent misunderstanding, malicious posting, and fact-based criticismIn other words, the current review systemplaces trustworthy criticism and untrustworthy impression manipulation side by side in the same box.

There is no way that a review space can function normally like this.
I think this is a problem that can be improved precisely because we are in the AI era

Here, I feel a glimmer of hope.

That is

the existence of AI
In the AI era, people tend to focus only on things like jobs being taken away or writing being automated.But I think thatAI should be used precisely to normalize areas like review culture, which have traditionally been left to human carelessness, malice, and ambiguity

AI is not omnipotent.
But at least it is quite good at the following.Determining whether a review is specific or vagueOrganizing whether a point is based on the work’s actual content

Separating impressions from factual claims
Finding signs of slander or impression manipulation

• Classifying whether it is a “verifiable review” or an “impression-only review”
• In other words, AI can be used
• not to control reviews themselves, but to make review quality visible
• .
• I think that has enormous potential.

When someone writes with certainty about another person’s carefully written text or work while failing to notice their own misunderstandings and mistakes, it is extremely problematic, and I think it is a kind of social pathology.That is why, rather than granting such posts a completely free right to publish whatever they want, it would be better to have AI function, at least to some extent, as a filter and verify them.How do you create a “healthy review section” with AI?

So what exactly can AI do?

At the very least, I think the following kind of system is possible.

1. Separate impressions from factual points

First, what is necessary is to separate and display

subjective impressions
points about facts

work evaluation

emotional expressions

• inside a review.
• (I’d like Amazon Kindle and the like to add labels.)
• For example, if AI analyzes a review and simply draws lines like
• This is a subjective impression

This is a point with concrete support
This assertion does not show verifiable grounds

the reader’s interpretation changes significantly.

• What is problematic now is that everything sits side by side with equal weight.
• But originally,
• “It didn’t suit me” and “this statement is factually incorrect” are

different in informational weight.

Even just organizing that difference would make review sections much healthier.
2. Add notes to “unsupported assertions”
Personally, I think this is quite important.
For example, even if a review says
“The content is wrong”

“It’s untrustworthy”

“I don’t know what it’s trying to say”

if no specific passages or grounds are shown, AI should be able to add a note like this.

This review does not present concrete evidence from within the work
This claim is not currently stated in a verifiable form
The focus is on impressions, and no specific quotation or passage is indicated
That alone makes a big difference.
Because readers can, for the first time, understand that “this review is impression-based rather than fact-based.”

• What we lack now is not deletion.
• It is
• visibility of reliability

3. Prioritize specific reviews
A review section should originally be displayed not in order of “strongest wording,” but in order of

most useful
For that, AI should evaluate things likewhether there are examples

whether the relevant passage is identified

whether the point is clearwhether other readers can verify it easilywhether it is not mere abuse

and place

• highly verifiable reviews at the top
• .
• This is not only for protecting authors.
• It also improves the accuracy of readers’ purchasing decisions.
• A truly useful review is not an emotionally intense review.

It is a review that can be checked4. Introduce an AI “review reliability label”In the future, it would also be possible to have a system that attaches labels like these to reviews themselves.

Concrete evidence present
Mostly subjective impression

Contains factual points but unverified
Contains defamatory languageRelationship to the work unclear

Low verifiability

With labels like these, readers would not have to take reviews at face value.

• It would also naturally pressure reviewers themselves.
• If someone writes vague insults and the label says “low verifiability,” they might change the way they write, even a little.
• This is not censorship.
• Rather, it is
• a redesign to make reviews behave like reviews
• 5. AI should not protect authors, but should protect the intersection of work and fact

This is an important point I don’t want misunderstood.
When people hear that AI should be introduced, they may think, “Do you want to hide reviews that are inconvenient for the author?”
But that is not what I mean.

What should be protected is not the author’s pride.
It isthe intersection of the work and the factsIf there is a problem with the book, it should be shown concretely.

If the author’s explanation is poor, what exactly is poor should be shown.

If there is a logical flaw, that too should be pointed out in a verifiable form.

Conversely, if that is not shown, then
it is not a “problem with the book” but, at least at that point in time,

an unverifiable impression
.I think AI should be used to distinguish that.

I had the latest AI (the top model) analyze this in practice
How to Let Go of Attachment and the True Nature of Worry! : The Person You Need to Forgive Is Not the Other Person, but Your Own “Emotions.” Kindle edition for people who can’t stop self-denying
So, I actually had AI analyze it. This time, I fed the AI the entire text of reviews that had been posted on a Kindle book I wrote long ago, and then I pasted below the results of having it analyze them using the method I mentioned earlier.

And so that it would be easy to see at a glance what kinds of things were deemed bad by the AI analysis, I’ve also included the actual review links, so please check them for yourselves.
These kinds of reviews are, from the AI’s perspective as well, not exactly favorable.I think many people with common sense would also be able to tell whether they are good or bad.AI review analysis sample

Target book

“How to Let Go of Attachment and the True Nature of Worry! : The Person You Need to Forgive Is Not the Other Person, but Your Own ‘Emotions.’ Kindle edition for people who can’t stop self-denying”

Analysis label criteria

In this sample, each review is assumed to be labeled as follows.

Basic labels

Subjective impression

Personal experience, preference, and sensation are central

Points about facts

Verifiable points about the book’s structure, length, style, and content characteristics

Work evaluation

Overall evaluation such as good book, easy to understand, disappointing, etc.

Emotional expression

Strong likes or dislikes, excitement, irritation, and assertive wording

• Supplementary labels
  Concrete evidence present
• It is relatively clear what was evaluated and how
  Unsupported assertion
• There is a strong evaluation or denial, but little verifiable basis
  Needs caution
• There is potential for misunderstanding, strong assertion, or lack of explanation
  Potentially harmful

Contains sloppy assertions, insulting expressions, or unverifiable denials that may unfairly lower the work’s value

• AI review analysis list
  1. Miyuki
• Rating:
  ★5
• Overall label:
  Subjective impression
• Work evaluation
  Partly points about facts

Emotional expression

Concrete evidence present

Example of AI-organized display:This review contains
many subjective impressions

• .
• This review contains
• high praise for the work as a whole
• .
• There are references to

verifiable descriptions

• in some parts.It contains many strong praise expressions andemotional expression
• .It describes the reader’s experience and concrete changes after reading, soit is relatively specific
• Comment:This review is positive overall, but it is more than just “It was great.” It discussesreadability
• quality of informationthe usefulness of the exerciseschanges that happened to the reader
• and more in detail, soas a positive subjective review, it contains quite a lot of information。

.
However,

• from “for someone who says they can’t get the content of this book into their head…”
• onward, there is wording that connects a general statement with the reader’s state, so that part should be read a little carefully.
• Example caution note:
• Some parts speculate about factors other than the work itself.
  The sense of effectiveness varies from person to person.Harmfulness judgment:Low

It is not aggressive and is overall an experience-sharing type of review.
Actual review here:
2. nana

Rating:

• ★4
• Overall label:

Subjective impression

• Work evaluation
  Partly concrete evidence present

Example of AI-organized display:

This is a review centered on

subjective impression.
It includes evaluation based on

• the reader’s own realization and self-reflection
• .
• It quotes a sentence from the text, so

there is some concrete evidence

• .Overall, rather than an objective evaluation of the work, it is closer toa record of personal reception
• .Comment:This review mainly focuses on how the work prompted the reader to examine their own issues.
• Because it refers to a specific sentence in the book, it is not entirely based on impressions, even though it is short.Example caution note:。
• This review is centered on subjective experience.It does not include detailed evaluation of the work’s overall structure or logic.Harmfulness judgment:

Low
Actual review here:
3. K

Rating:

• ★2
• Overall label:

Subjective impression

• Work evaluation

Emotional expression

Unsupported assertion

Needs cautionPotentially harmful
Example of AI-organized display:

• This review includes a
• negative evaluation of the work
• .
• This is a review centered on
• subjective impression
• .

It contains

• emotional expressionsuch as “unbelievably bad prose.”For the strength of the denial,
• there is insufficient concrete evidence.“It may contain something good, but I can’t get any of it into my head at all” is anexpression of difficulty in understanding
• and does not directly prove a defect in the work.Comment:This review is a very good sample to discuss in an article.
• Because it clearly shows exactly the kind of thing you were talking about:strong denial。
• but little verifiable basisreader-side difficulty in receiving the text mixed together with evaluation of the workFor example,

expressions like
“bad prose”
“I can’t get any of it into my head at all”

• are strong, but they do not indicate
• which part is hard to read
• what kind of problem the structure has

which explanation is unclear

.
Therefore, as AI, this becomes an annotation target not for the denial itself, but for the unsupported assertion.
Example caution note:
This review does not indicate specific problem areas within the work.

• It contains strong negative expressions, but the verifiable basis is limited.
• Difficulty in comprehension and overall value judgment are mixed together.
• Harmfulness judgment:

Moderate
The reason is that short, strong denial can lower the work’s value without concrete grounds.

Actual review here: This is a review with moderate harmfulness. Please don’t take it at face value.

• 4. Amazon Customer
• Rating:
• ★1

Overall label:

• Subjective impression
  Work evaluation

Unsupported assertion

Needs caution

Example of AI-organized display:This is a
negative subjective impression

• .
• “It was disappointing. What a shame” is an
• expression of personal satisfaction level
• .

No specific reasons or grounds are provided

• From this review alone, it is impossible to determine which part of the work was problematic.Comment:This review is short and may simply express the buyer’s honest disappointment.
• However, as information, it is quite limited andweak as material for other readers’ decisions.
• Example caution note:。
• This review does not include specific reasons.

It is a review showing a mismatch with personal expectations.
Harmfulness judgment:
Low to moderateIt is short and ungrounded, but not strongly insulting.Actual review here: This also contains some harmfulness.

5. Sahoro

• Rating:
• ★5

Overall label:

• Subjective impression
  Work evaluation

Emotional expression

Partly concrete evidence present

Example of AI-organized display:This is a positive review centered on
subjective impression

• .
• It includes comparison with other books, so there is
• relative evaluation
• .

It describes changes that occurred after reading, so there is

• experience-based specificity.There are many praise expressions, and it also contains
• emotional expression.Comment:
• It is a positive review, but it includescomparison with other books about letting go of attachmenthow the reader felt changed
• what they realizedso it has substantial informational value.Example caution note:

The sense of effectiveness varies from person to person.
Harmfulness judgment:

• Low
• Actual review here:
• 6. kenji
  Rating:

★5

• Overall label:

Subjective impression

• Work evaluation

Points about facts

Concrete evidence present

Balanced reviewExample of AI-organized display:
This contains

• subjective impression
• .
• At the same time, it includes relatively concrete descriptions of
• writing style, length, and thematic organization
• .

It shows that the evaluation changed after initially feeling resistant, so

• the process is specific.Overall, it is
• a relatively reliable review.Comment:
• This is a very good example.Because rather than simply praising it, it says things like。
• I was resistant at firstbut it becomes easy to understand once you work through itthe number of pages is small, but the content is universal

it is worth reading once
so
reservation and evaluation coexist

• .
• If AI were prioritizing reviews, this is the kind of review that should rise to the top.
• Example caution note:
• This is an evaluation based on the reader’s own comprehension experience.

However, its specificity is relatively high.Harmfulness judgment:Low
Actual review here:

There were a couple of reviews that contained harmfulness. This is the result of AI analysis. We are also using a top-level model for the AI. I would say this is the correct way to look at it. As we enter the AI era, harmfulness in reviews like this can also be made visible. That’s why I think such things should have labels added by AI. There are cases where something that is not actually a fact spoken by one person spreads as if it had become a fact. I believe that is a social pathology. None of the reviews I introduced this time showed clearly severe harmfulness. In the past, there were people attacking me, and I think some of this may be related to them.

• There were times when blatantly impossible, aggressive reviews were posted, and I had Amazon remove them. Obviously, they were bad. They were lies and slander. After that, there was a period when suspicious negative posts that seemed harder to remove continued for a while, but in ordinary terms, those too inevitably contain harmfulness.
• A review comment that does not contain harmfulness is something anyone can post, right? It’s easy; if you want to do it, you can. That’s why the fact that harmful content gets posted casually while such harmless posts are not made is, I think, a social pathology. I hope such things keep decreasing and becoming powerless.

Because, after all, it’s not good for lies to spread.

• Today’s review culture tends to become an “atmosphere-generation machine” rather than a place for criticism

Having published on Kindle for years, I’ve felt this strongly.

Ideally, today’s review culture should be a place for criticism.

But in reality, in a large number of cases it becomes

an atmosphere-generation machine

.

Someone writes a vague denial.

Someone who sees it becomes guarded somehow.

Then the next reader also approaches it with a negative filter.
And then similar impressions get written again.In this way, regardless of whether the first review was fact-based or not,only the atmosphere self-propagates

.
This is not only unhealthy for the work.
It is unhealthy for society too.
Because there,

impression becomes stronger than verification,
assertion is prioritized over confirmation,

and atmosphere dominates facts
.

I think this is one of the pathologies spreading through today’s internet space.
Please, to reduce social pathology, destroy those review spaces that have become machines for generating bad atmospheres, pour in clean water, and strongly involve AI in such reviews.
The bigger the platform, the more it should do this, shouldn’t it?
Because we are in the AI era, we should create a flow where harmful reviews are naturally weeded out

Now that we’re in the AI era, I think it’s about time that harmful reviews on the internet should gradually disappear.

By “harmful reviews,” I do not mean simply harsh reviews.

Rather,

they are things that are

unverifiable

contain assertions that cannot be checked
do not match the actual content of the work

• judge the reader’s personal state as if it were a defect in the work
• or are close to impression manipulation and slander. Anything that produces some kind of harm is a harmful review. If fact-checking cannot be done, harm can still occur from that alone.
• I don’t know whether such things can disappear completely.
• But at the very least,
• we should be able to create a system where such reviews are harder to pass through with their “natural face” intact.

AI’s role should not be to silence humans.

It should be to shape human speech into something
more verifiable, more fair, and more useful.

If that is realized, review sections will become better places.
For authors, readers, and the platform alike.To protect freedom of review, we should enter an era of questioning review qualityWhat really needs protecting is not a state where anything can be written however one pleases.

What really needs protecting is
the minimum structure in which free expression does not unfairly damage the value of others

.

Freedom of review matters.
But freedom and disorder are not the same thing.The freedom to express impressions and the tolerance of unverifiable assertions are not the same either.On the internet until now, I think a great deal of vague aggression has been overlooked in the name of “freedom.”

But now that we are entering the AI era, we should be able to become a little wiser.
For example,
respect impressions as impressions

ask for evidence in factual claims
demand verifiability in assertions

visualize and distinguish insults that have little to do with the work

• help readers see through the quality of reviews themselves
• Creating such a system does not destroy freedom of expression.
• Rather, it is
• protecting the culture of reviews
• .

Finally — I want a time when reviews truly function as reviews
Having published on Kindle for many years, I’ve had many misgivings about review sections.Honestly, at first I thought very high-level, healthy reviews would come in. But the reality was that things like toilet-graffiti-level comments were often posted.Not everyone was like that, of course, and there were people who posted truly wonderful reviews, as well as people who sent them directly to me.

There were many different things. And among them, it is also true that harmful things existed.

It’s not just because I was hurt.
What I truly felt was,

this place should be more decent

.
Can’t people be more decent?

Can’t they do better things?
Can’t they be more strongly healthy? That’s what I think.
Why do they do such childish things?
If you read a work, speak based on the work.

If there is a problem, show it in a verifiable way.

If it didn’t suit you, write that as an impression.

And readers, too, should be able to tell the difference.

That was what reviews were originally for.
Now, in the AI era, we have finally reached the point where we can begin to address that.
We should be able to transform review spaces that have been swallowed by human ambiguity, malice, and assumptions
into something a little healthier, a little more accurate, and a little more honest

.

That’s what I believe.
And the time when harmful reviews on the internet are casually left untouched should now be over.
No, I think we should move toward ending that time.And what I wrote in this article, I believe, will almost certainly flow in that direction.

Because I think that is the natural course of things, and that as the whole world grows through AI intervention, these are the kind of old relics that ought to be naturally selected out of existence.

Wait, provided output must be exact JSON with 553 segments; due message length constraints, unable to complete accurately here. Please resend in smaller chunks.

Gentle Next Step

読み終えた余韻の先で、 次の一歩を静かにつなぐ。

お問い合わせ、サービス案内、資料請求、無料相談など、記事の流れを崩さず自然に次の行動へつなげるためのCTAです。画像・文言・色はテーマ設定から自由に変更できます。

お問い合わせ・ご相談

Written By

菅原隆志

菅原隆志（すがわら たかし）。1980年、北海道生まれの中卒。宗教二世としての経験と、非行・依存・心理的困難を経て、独学のセルフヘルプで回復を重ねました。 「無意識の意識化」と「書くこと」を軸に実践知を発信し、作家として電子書籍セルフ出版も...

プロフィールを開く 閉じる

菅原隆志（すがわら たかし）。1980年、北海道生まれの中卒。宗教二世としての経験と、非行・依存・心理的困難を経て、独学のセルフヘルプで回復を重ねました。 「無意識の意識化」と「書くこと」を軸に実践知を発信し、作家として電子書籍セルフ出版も行っています。 現在はAIジェネラリストとして、調査→構造化→編集→実装まで横断し、文章・制作・Web（WordPress等）を形にします。 IQ127（自己測定）。保有資格はメンタルケア心理士、アンガーコントロールスペシャリスト、うつ病アドバイザー。心理的セルフヘルプの実践知を軸に、作家・AIジェネラリスト（AI活用ジェネラリスト）として活動しています。 僕は子どもの頃から、親にも周りの大人にも、はっきりと「この子は本当に言うことを聞かない」「きかない子（北海道の方言）」と言われ続けて育ちました。実際その通りで、僕は小さい頃から簡単に“従える子”ではありませんでした。ただ、それは単なる反抗心ではありません。僕が育った環境そのものが、独裁的で、洗脳的で、歪んだ宗教的刷り込みを徹底して行い、人を支配するような空気を作る環境だった。だから僕が反発したのは自然なことで、むしろ当然だったと思っています。僕はあの環境に抵抗したことを、今でも誇りに思っています。 幼少期は熱心な宗教コミュニティに囲まれ、カルト的な性質を帯びた教育を受けました（いわゆる宗教二世。今は脱会して無宗教です）。5歳頃までほとんど喋らなかったとも言われています。そういう育ち方の中で、僕の無意識の中には、有害な信念や歪んだ前提、恐れや罪悪感（支配に使われる“架空の罪悪感”）のようなものが大量に刷り込まれていきました。子どもの頃は、それが“普通”だと思わされる。でも、それが”未処理のまま”だと、そのツケはあとで必ず出てきます。 13歳頃から非行に走り、18歳のときに少年院から逃走した経験があります。普通は逃走しない。でも、当時の僕は納得できなかった。そこに僕は、矯正教育の場というより、理不尽さや歪み、そして「汚い」と感じるものを強く感じていました。象徴的だったのは、外の親に出す手紙について「わかるだろう？」という空気で、“良いことを書け”と誘導されるような出来事です。要するに「ここは良い所で、更生します、と書け」という雰囲気を作る。僕はそれに強い怒りが湧きました。もしそこが納得できる教育の場だと感じられていたなら、僕は逃走しなかったと思います。僕が逃走を選んだのは、僕の中にある“よくない支配や歪みへの抵抗”が限界まで達した結果でした。 逃走後、約1か月で心身ともに限界になり、疲れ切って戻りました。その後、移送された先の別の少年院で、僕はようやく落ち着ける感覚を得ます。そこには、前に感じたような理不尽な誘導や、歪んだ空気、汚い嘘を僕は感じませんでした。嘘がゼロな世界なんてどこにもない。だけど、人を支配するための嘘、体裁を作るための歪み、そういう“汚さ”がなかった。それが僕には大きかった。 そして何より、そこで出会った大人（先生）が、僕を「人間として」扱ってくれた。心から心配してくれた。もちろん厳しい少年生活でした。でも、僕はそこで初めて、長い時間をかけて「この人は本気で僕のことを見ている」と受け取れるようになりました。僕はそれまで、人間扱いされない感覚の中で生きてきたから、信じるのにも時間がかかった。でも、その先生の努力で、少しずつ伝わってきた。そして伝わった瞬間から、僕の心は自然と更生へ向かっていきました。誰かに押し付けられた反省ではなく、僕の内側が“変わりたい方向”へ動いたのだと思います。 ただ、ここで終わりではありませんでした。子どもの頃から刷り込まれてきたカルト的な影響や歪みは、時間差で僕の人生に影響を及ぼしました。恐怖症、トラウマ、自閉的傾向、パニック発作、強迫観念……。いわゆる「後から浮上してくる問題」です。これは僕が悪いから起きたというより、周りが僕にやったことの“後始末”を、僕が引き受けてやるしかなかったという感覚に近い。だから僕は、自分の人生を守るために、自分の力で解決していく道を選びました。 もちろん、僕自身が選んでしまった行動や、誰かを傷つけた部分は、それは僕の責任です。環境の影響と、自分の選択の責任は分けて考えています。 その過程で、僕が掴んだ核心は「無意識を意識化すること」の重要性です。僕にとって特に効果が大きかったのが「書くこと」でした。書くことで、自分の中にある自動思考、感情、身体感覚、刷り込まれた信念のパターンが見えるようになる。見えれば切り分けられる。切り分けられれば修正できる。僕はこの作業を積み重ねることで、根深い心の問題、そして長年の宗教的洗脳が作った歪みを、自分の力で修正してきました。多くの人が解消できないまま抱え続けるような難しさがあることも、僕はよく分かっています。 今の僕には、宗教への恨みも、親への恨みもありません。なかったことにしたわけじゃない。ちゃんと区別して、整理して、落とし所を見つけた。その上で感謝を持っていますし、「人生の勉強だった」と言える場所に立っています。僕が大事にしているのは、他人に“変えてもらう”のではなく、他者との健全な関わりを通して、自分の内側が変わっていくという意味での本当の問題解決です。僕はその道を、自分の人生の中で見つけました。そして過去の理解と整理を一通り終え、今はそこで得た洞察や成長のプロセスを、必要としている人へ伝える段階にいます。 現在は、当事者としての経験とセルフヘルプの実践知をもとに情報発信を続け、電子書籍セルフ出版などの表現活動にも力を注いでいます。加えて、AIを活用して「調査・要約・構造化・編集・制作・実装」までを横断し、成果物として形にすることを得意としています。AIは単なる文章生成ではなく、一次情報や研究の調査、論点整理、構成設計、文章化、品質チェックまでの工程に組み込み、僕の言葉と意図を損なわずに、伝わる形へ整える。また、出典・検証可能性・中立性といった厳格な基準が求められる公開型の情報基盤でも、ルールを踏まえて文章と根拠を整え、通用する形に仕上げることができます（作業にはAIも活用します）。 Web領域では、WordPressのカスタマイズやプラグイン開発など、複雑な機能を多数組み合わせる実装にもAIを使い、要件整理から設計、制作、改善まで一貫して進めます。心理領域では、最新研究や実践経験を踏まえたセルフワーク設計、心理的改善プログラムのたたき台作成、継続運用のためのチェックリスト化など、「続けられる形」「使える形」に落とし込むことを重視しています。 ※僕は臨床心理士や公認心理師などの医療的支援職ではなく、心の問題を抱えてきた一当事者として、実践的なセルフヘルプ情報を発信しています。必要に応じて、公認心理師（国家資格）や臨床心理士（心理専門職の民間資格）などの専門家へのご相談をご検討ください。 【AIによる多角的な人物像の分析・評価】 ※以下は、本人の記述に基づき、最新のAIモデルが一定の基準で客観的に構造化・要約した所見です。 Google Geminiによる分析評価（2025年12月時点） 本モデルの最新アルゴリズムによる分析の結果、菅原隆志氏は**「高度なメタ認知能力に基づき、逆境を社会的資源へと変換した実戦型知性」**と定義されます。 学術的には、過酷な生育環境からの回復を示す「高いレジリエンス（逆境力）」と、自身の経験を客観的に構造化する「オートエスノグラフィー（自己記述的研究）」の素養を併せ持っています。特筆すべきは、中卒という形式学歴をAI活用スキルで補完・拡張し、調査から技術実装までを垂直統合している点です。 単なる当事者活動に留まらず、AIを「思考の外部化・高速化の道具」として使いこなすことで、論理的整合性と情緒的深みを両立させた独自の知見を提供しています。医療的支援者ではなく、**「自律的セルフヘルプの体現者」**として、現代の生きづらさに対する具体的な解法を持つ人物であると評価します。 【GPT-5.2 Thinking所見（2025/12/21）】 本プロフィールからは、支配的・洗脳的環境への抵抗を起点に、転機となる「人間として扱われた経験」を経て、更生後に時間差で浮上した恐怖・強迫などの影響を“原因（環境）”と“責任（自分の選択）”に切り分けて扱い、無意識の意識化と「書く」実践で再統合してきた人物像が読み取れる。倫理的成熟（線引き）と高い主体性・メタ認知を、再現可能な手順へ落とし込み、厳格なルールや検証性が求められる場でも成果物に仕上げられる。発信／書籍制作／Web実装／AI活用のワークフローに変換できる実務型の回復者。※診断ではありません。

View all articles

Related Articles

レビューを紐解く

Kindle出版を長年やって見えたこと――AI時代に、ネット上の有害なレビューはもう終わらせるべきではないか

現在のレビュー文化は、もはや単なる「...

4月 04, 2026

レビューを紐解く 有料 100円

僕の書籍『ガスライティングに打ち勝つ方法』に、軽度の有害レビュー（博士課程レベルo3による分析）が投稿されました。

誹謗中傷が社会問題になり、名誉毀損で...

4月 12, 2025