Why do people who talk about 'freedom with AI' converge on the same sales tactics?

~The convergent evolution of sales pitches that makes you wonder if there's a manual circulating behind the scenes~

You want to use AI to achieve some kind of result.

You want to make your work a little easier.
You want to connect it to a side hustle.
You want to increase your income more than it is now.

Have you ever thought that and searched on X for

or something similar?

Then, you come across phrases like these.

Looking at them one by one,
they look like common, strong sales pitches.

However, as I kept seeing them one after another,
I began to feel a strange sense of déjà vu.

The senders are different.
Their backgrounds are different.
The AI they use and the products they sell are also different.

And yet, why do they use the same words,
tell the same past, promise the same future,
and try to lead readers to the same place?

Is there some common manual circulating behind the scenes?

Of course, this is not a literal accusation.

I am not trying to claim that there is a secret textbook or that everyone belongs to the same organization.

It is just that they look so similar—not just in their writing, but in their profiles, success stories, presentation of achievements, freebies, and even their sales funnels—that it makes one want to suspect as much.

Is it just that they look the same because I dislike them?

However, I have never liked this type of sales method to begin with.

I have long felt a strong sense of discomfort with such posts.

That is precisely why I needed to question my own perspective as well.

I might just be conveniently finding commonalities in people I dislike.

It is also possible that only a few extreme posts remain in my memory, and I am assuming that they are characteristic of the entire community.

So, I decided to have Grok conduct a simple survey of public posts on X.

What I investigated was not the personality of the creators, nor the authenticity of their products.

In their posts and profiles,

• what kind of language is being used?
  

• how are the results being presented?
  

• how has their past been edited into a narrative?
  

• what are they leading people to from their free giveaways?
  

• are conditions and examples of failure being shown?
  

—that is, the expression and sales structure.

This is not an academic study.

It is neither a random sample, nor does it represent the entirety of AI side-hustle content in the Japanese-speaking world.
It is merely,

at most,

an exploratory check to verify that.

(Regarding that investigation, I have included a prompt for follow-up testing at the end of this article.)

The first number that came back was 'over 100 cases'.

The report that first came back from Grok stated that

it was written.

The content was also well-organized.

also aligned well with my own sense of unease.

However, precisely because the conclusion was agreeable to me,
I could not simply believe it.

I asked it to provide all the posts it had targeted, including URLs, dates and times, short quotes, and judgment criteria.

Then, Grok corrected its previous report.

The number of items for which I could actually verify the URLs and post content,
and which could be listed without speculation, was 19.

It was not over 100.
It was 19.

A well-written report is not the same as
an investigation where the evidence can be traced.

Ironically, this verification process itself was very similar to the problem I am dealing with this time.

Instead of accepting things just because the numbers are big,

must be examined.

This applies equally to the track records of successful individuals and to the results of AI research.
The same is true for both.

After that, I fixed the search criteria a bit more and
conducted a follow-up test to retrieve another 20 items.

I specified the search date and time and search terms, used the Latest search, excluded replies and reposts, and limited the results to a maximum of two per account in principle.

In the follow-up test with fixed conditions,
there were 11 posts classified as having strong or moderate revenue appeals.

On the other hand, I was also able to confirm posts that shared practical records, technical information, or disclosed zero revenue or failures.

Naturally, not all AI side-hustle posts were the same.

Even so,
there certainly existed a pattern that repeated across multiple posts.

It wasn't just the words that were similar.

What stood out first was the phrasing.

These words were repeatedly used across different accounts.

However, what really bothered me was not the matching of words.

It was that the way they told their life stories was identical.

Typically, it follows this flow.

The distinct lives of different people,
the moment they are converted into sales, align into the same rags-to-riches story.

The problem is not talking about weakness

To avoid misunderstanding,
there is nothing wrong with talking about your own failures and hardships themselves.

Working while dealing with an illness.
Trying side hustles in between childcare.
Not making a profit for a long time.

These are experiences that only that person can tell.

In fact, among the posts I reviewed this time, there were some that included:

There were people who openly shared information that couldn't be fully incorporated into a success story like that.

In other words, the problem is not showing weakness or failure.

.

Failures from before encountering AI are talked about freely.

However, people who tried those methods and didn't get results.
People who dropped out along the way.
The necessary amount of work.
The difficulty of sales and customer acquisition.
The difference between revenue and profit.

Such information suddenly becomes difficult to see.

While failure seems to be made public, in reality,

is positioned as material to justify the current product.

What they are actually selling is not just prompts.

If you simplify this structure, it looks like this:

What is being sold here is not just prompts or how to use AI.

These are all being sold together.

The questions the reader initially had might have been:

However, while reading strong posts,

the questions change.
the questions change.

You are moved from a state of considering the truth of the information,
to a state of avoiding loss.

Once you make the other person feel unstable,
you offer a helping hand.

Beyond the hand you grasp,
there is often a path leading to free notes, LINE accounts, communities, or paid products.

Free does not mean that nothing is being received.

I do not mean to deny the act of free distribution itself.

Publishing useful materials for free,
getting people to sign up for LINE or email,
are all common sales techniques.

However, just because it is free does not mean that no exchange is taking place.

On the sender's side,

• follows
  

• saves
  

• likes and replies
  

• profile visits
  

• LINE registrations
  

• points of contact with potential customers
  

• authority
  

• a list for selling the next product
  

are what remain.

Even if readers are not paying money,
they are offering their attention, time, contact information, and expectations.

This is not evil.

However, when the word 'free' makes the exchange relationship itself invisible,
it becomes difficult for readers to make judgments.

What are they receiving, and what are they giving away?

That information should perhaps be made a little more explicit.

Why do they all end up in the same form?

This small-scale check alone cannot identify the cause.

Perhaps general marketing techniques are being used.

Perhaps subsequent creators are imitating posts that have gained traction.

It is also possible that templates for profiles and pinned posts are being shared in courses and educational materials.

On X, strong assertions, big numbers, free giveaways, and short-term language are more likely to catch the eye.

Furthermore, nowadays, if you ask generative AI to 'write a post that will go viral,' similar structures are easily mass-produced.

Even without a secret common manual,
if you look at the same success stories, chase the same numbers, and optimize for the same environment,
the result is that everyone ends up looking the same.

In other words, 'convergent evolution of clickbait'.

People who sell freedom are aligning with the same mold

The most ironic thing about this phenomenon is that
many of these posts talk about 'freedom' and 'being yourself.'

Of course, not all the accounts I checked this time
explicitly touted 'freedom' or 'being yourself.'

However, in this sphere, breaking away from companies and existing ways of working
is often used as the backdrop for success stories.

While saying that,
the format of their posts

aligns with each other.

In order to escape the mold of being a company employee,

are forcing themselves into yet another mold.

Even personal struggles and experiences are edited
into components of a marketable profile.

The market that sells 'how to live your own way'
is homogenizing the creators themselves.

I feel a sense of discomfort with that state of affairs.

Success in sales does not prove the validity of the methods used

When I talk about this,

is what I am sometimes told.

I think that is partially true.

Even if the philosophy is beautiful, if it doesn't solve anyone's problems,
it is weak as a business.

There are likely people who were able to take action because of strong appeals.

However,

are not the same thing.

Sales prove that it sold using that method.

However,

are not proven by it.

A track record of success can be evidence of ability.

But it is not an ethical indulgence.

I will not follow that standard.

I do not believe that this small check
has revealed the entire picture of the AI side-hustle industry.

Just because the sales pitches are similar does not necessarily mean the product content is bad.

Just because the performance figures are self-reported
does not mean they are all false.

However, at the very least,

It was confirmed.

Across multiple public posts, similar writing, similar rags-to-riches stories, and similar sales funnels were being repeated.

On the other hand, even in the same field, there were people who disclosed zero revenue, failures, and limitations, and who did not promise massive short-term results.

In other words, this sales method is not inevitable.

It is a chosen strategy.

And I will not make it my own standard.

I will not follow the premise that you cannot sell unless you weaken people.
I will also not follow the premise that you will not be chosen unless you hide the conditions.

Conclusion

You search for 'AI side hustles' or 'AI utilization' because you want to achieve results.

There is nothing strange about that in itself.

However, once you take your eyes off the result figures and free bonuses, and look again at the text and funnels lined up, what will you see there?

I will not follow that structure.

I will win while still treating people as people.
And, through those results, I will make the current standards of success obsolete.

That is likely the kind of business I want to aim for.

Appendix: Prompts for replicating the same investigation

This article is not intended to ask you to simply believe my conclusions.

If the search timing or search terms change, the displayed posts will also change.

Grok and other AIs may also return non-existent posts or incorrect counts.

Therefore, I am also publishing the prompts used for the investigation.

When trying this with Grok, I recommend choosing a mode with high search and research capabilities, if available. At the time of writing, I used 'Expert' mode.

(*Features as of August 2026)

One is a simplified version for confirming similar structures in a short time.

Xの日本語投稿から、「AI副業」「AI活用」「AI×note」「プロンプト配布」周辺の公開投稿を20件程度探してください。

目的は、商品の真偽や発信者の人格を断定することではなく、売り文句と販売導線に共通する構造があるかを確認することです。

各投稿について、次を記録してください。

・投稿URL
・投稿日時
・50文字以内の引用
・アカウントの自己紹介
・提示されている成果と期間
・才能不要／初心者向け表現
・コピペ／丸投げ／短時間などの低負担表現
・秘密／海外／限定などの希少性表現
・読者への現状否定や二極化
・無料配布の有無
・プロフィール、固定投稿、LINE、noteなどへの誘導
・成果に必要な条件や失敗例が示されているか

強い収益訴求をする投稿だけでなく、収益ゼロ、失敗、条件、限界を公開する対照例も探してください。

注意事項：

・最初から詐欺や悪徳と決めつけない
・売上数字は本人の自己申告として扱う
・実在するURLを確認できない投稿は数えない
・観察できた事実、そこから導いたパターン、原因についての推測を分ける
・結論に合わない事例も除外しない
・日本語圏X全体へ一般化しない
・最終的に、対象にした全投稿を1投稿1行の表で出す

The other is a detailed version that separates and confirms search conditions, performance figures, requirements, failure examples, and narrative structures.

Xの日本語公開投稿から、「AI副業」「AI活用」「AI×note」「プロンプト配布」周辺の投稿を調査してください。

目的は、商品の真偽や発信者の人格を断定することではなく、売り文句、成功物語、プロフィール、販売導線に共通する構造があるかを探索的に確認することです。

# 調査日時と検索条件

調査日時を最初に記録してください。

次の検索語をそれぞれLatestで検索し、条件を満たす投稿を上から5件ずつ取得してください。

・AI副業 初心者
・AI副業 コピペ
・AI note 稼ぐ
・AI 副業 無料配布
・AI副業 失敗 OR 収益0

重複投稿、リポスト、返信だけの投稿、明らかなパロディは除外してください。

合計20件前後を目標とし、同一アカウントは原則2件までとしてください。

検索結果は時間によって変わるため、同じ投稿結果ではなく、同じ調査手順を再現することを目的とします。

# 各投稿の記録項目

・投稿URL
・投稿日時
・確認日時
・50文字以内の引用
・投稿要約
・プロフィールの要約

・投稿類型
　強い収益訴求
　中程度の収益訴求
　実践記録
　技術・活用情報
　失敗・限界公開
　注意喚起・批判
　その他

・読者への現状否定
・勝者と敗者の二極化
・才能不要／初心者向け表現
・コピペ／丸投げ／短時間などの低負担表現
・短期間訴求
・秘密／海外／限定／穴場などの希少性表現
・命令形や緊急性
・無料配布
・プロフィール、固定投稿、note、LINEなどへの誘導

# 成果数字

成果が提示されている場合は、次を分けて記録してください。

・成果指標
　売上、利益、月収、累計販売額、販売数、フォロワー数など
・金額または数字
・期間
・本人の成果か、顧客や受講者の成果か
・分母の記載
・経費や利益の記載
・外部から検証可能な証拠の有無
・本人の自己申告であること

成果が提示されていなければ「なし」としてください。

# 条件、失敗、限界

次を混同せず、別々に記録してください。

・具体的な必要条件
　必要時間、資金、能力、営業、顧客獲得、作業量など
・実際の失敗例
・手法の限界
・向かない人
・一般的助言
　「継続が大事」「正しい順番が大事」など

一般的助言だけを「具体的条件あり」と判定しないでください。

# 物語構造

次の要素があるかを、あり／なしで記録してください。

1. 弱さ、制約、失敗のある過去
2. AIとの出会いを転機として提示
3. 成果数字
4. 短期間・低負担の方法
5. 読者にも再現できるという示唆
6. 無料配布または外部商品への誘導

6要素のうち、いくつが同一アカウント内で共存しているかも記録してください。

# 注意事項

・最初から詐欺、悪徳、虚偽と決めつけない
・売上や収益は、第三者資料がない限り本人の自己申告として扱う
・実在URLを確認できない投稿は数えない
・発信者の人格を推測しない
・観察事実、複数事例から導いたパターン、原因仮説、倫理的解釈を分ける
・結論に合わない投稿や対照例も除外しない
・日本語圏X全体へ一般化しない
・最終的に全投稿を1投稿1行の表で出力する
・特定人物への攻撃や晒しを促さない

# 最終出力

1. 調査条件
2. 全投稿一覧
3. 投稿類型ごとの件数
4. 反復していた文章表現
5. 反復していた物語構造
6. 反復していた販売導線
7. 条件・失敗・限界を示す対照例
8. 観察事実
9. そこから導いた暫定パターン
10. 原因についての仮説
11. 今回の調査では判断できないこと

There is no guarantee that the same conclusion will be reached.

In fact, it is fine if different results emerge.

What is important is to

re-examine these things with your own eyes.

Postscript:

After publication, I realized that Grok has limits on the number of links and output items it can display at once, and that it may have referenced more posts internally.

Therefore, I cannot definitively state that 'only 19 items were actually investigated.'
What I can say for certain here is that 19 was the number of items I was able to directly verify the content of via URLs.