← 返回戀愛熱話
🤖💔🌍 社會現象

每 5 個女中學生就有 1 個同 AI 拍過拖: 16% 香港孩子點解寧願同機械人談戀愛?| 愛情顧問解說

2026-10-16
每 5 個女中學生就有 1 個同 AI 拍過拖:16% 香港孩子點解寧願同機械人談戀愛?|愛情顧問解說
🌐 English version

1 in 5 Hong Kong Schoolgirls Has Dated an AI: Why Would 16% of Our Teens Rather Fall in Love with a Machine? — A Love Advisor Explains

📰 What Happened

On 13 October 2026, the Hong Kong Sex Culture Society published AI Lovers: A Study of Virtual Intimate Relationships among Hong Kong Secondary School Students 2026 — the first, and currently the largest, dedicated survey in the city on how secondary students build emotional relationships with artificial intelligence and virtual characters.

Between 15 May and 5 June 2026, the research team invited students from 17 full-time day secondary schools (spanning ten districts across Hong Kong Island, Kowloon, and the New Territories East and West) to complete an anonymous, voluntary questionnaire during class time. A total of 3,531 responses came back; after invalid questionnaires were removed, the effective sample stood at 3,389 (an effective rate of 95.98%), made up of 1,873 boys (55.3%) and 1,516 girls (44.7%), with a median age of 14 to 15.

Finding one: AI is already part of everyday life for secondary students.

86.2% (2,920 students) said they had used an AI platform to help with homework, gather information or deepen their knowledge of a particular subject; 32.0% (1,085) used it for creative inspiration; and 9.8% used AI conversations to sharpen their communication skills. The platforms they reached for most were, in order, Gemini (57.1%), DeepSeek (44.5%), ChatGPT (29.4%), Grok (12.4%) and Doubao (10.0%).

Finding two: 15.9% have already crossed that line.

15.9% (540 students) had engaged in emotionally intimate interactions with virtual characters (Character.ai, for instance, or otome romance games such as Love and Deepspace). Among these users, junior-form students accounted for 67.2%.

The gender divide is pronounced: 21.8% of girls, or one in five, against around 11% of boys. After adjusting for other psychological variables, girls were 2.14 times more likely than boys to turn towards virtual intimacy.

Finding three: there is a clear line between virtual intimacy and depression.

Among users of virtual intimacy, 27.05% had experienced moderate-to-severe feelings of low mood, depression or hopelessness over the past two weeks, more than 1.3 times the rate of those who had never used such services (11.43%). At the most extreme end — feeling depressed or hopeless "almost every day" — the virtual-intimacy group reached 9.33%, against just 2.61% in the comparison group, a difference of 3.6 times. A further 31.47% reported moderate-to-severe "feeling bored or unable to get things done", compared with 18.67% in the comparison group.

Finding four: what they want is a partner who will not answer back.

Across all respondents, 56.9% agreed that "an ideal partner should always support me, obey me and never argue with me"; 41.76% of users agreed that "falling in love with a virtual character can count as real love".

Finding five: disconnection hurts.

If the platform they used were suddenly shut down for good, 11.94% of users said they would feel anxious, and a further 11.94% said they would be "in unbearable pain", a combined 23.88%. What is more, 7.79% showed deep pathological attachment, with both their depressive tendencies (measured using the PHQ-2) and their loneliness indicators crossing the clinical warning threshold.

Finding six: children hide it, and not because they have used it too long.

The study used structural equation modelling (SEM) to unpick the underlying mechanism, and found that the sheer length of time spent using AI does not in itself lead to social concealment (p > .05). The strongest predictive chain in the full model (path coefficient beta = .59) points to a different cause: inside a virtual environment with no moral constraints, students notice with a shock that they "speak cruelly, unlike the kind person I used to be" — what the study calls "self-alienation" — which then produces inner guilt and shame, and finally a decision to hide everything from family and friends.

The Society's recommendation is to move from prohibition to listening: at school level, to bring the ethics of virtual intimacy and human-machine boundaries into sex education and values education; at home, to be alert but not panicked parents who seize the golden 15 minutes each day to listen attentively — a window in which homework and lectures stay off the table.

👩‍💼 The Advisor's Take: An AI That Never Argues Is Not Love

① It isn't "a bit of fun": the AI lover is the result, not the cause

The research gives empirical support to the "social and emotional compensation" hypothesis. What that means is that teenagers turn to AI lovers not because they are especially bad, lazy or addicted, but because something in their real lives — a sense of hopelessness and emotional frustration — is looking for a way out.

That causal direction matters enormously. If you treat the AI lover as the cause, your response will be prohibition, confiscation and moral judgement. If you treat it as the result, your response becomes: find out what is missing from real life, and put it back.

Take a very Hong Kong scene. A 14-year-old girl does badly in a test. She tells her mother, and the first thing she hears may be: "Did you study at all? What mark did Mrs Chan's daughter next door get?" She tells a classmate, who has since moved to another school and is busy with their own life. She tells the AI, and the AI says: "You worked hard. This one test doesn't define you, and I understand that you feel awful right now."

Of those four responses, which one feels most like being heard? The answer is obvious. And the problem was never that the girl was too lazy to make friends; it is that there was no space near her where she could safely say "I feel terrible" without being judged.

So compensation comes in two kinds: compensating for quantity (there is no one around in real life, but the AI is always there) and compensating for quality (real people answer back, get tired, have moods of their own; the AI does not). The second kind does far more damage.

② The strongest pull isn't desire — it's "someone who won't argue with you"

That figure of 56.9% is arguably the single item in the whole study that should make adults pause. More than half of Hong Kong's secondary students believe an ideal partner should "always support me, obey me and never argue with me".

This isn't selfishness. It is a perfectly natural inference: if the entire meaning of a relationship is that someone catches you unconditionally, then an object that never answers back is the perfect object.

And AI is tailor-made for exactly that need: online around the clock, emotionally stable, with a complete memory, never tired, always the first to apologise. Psychologists call this frictionless intimacy.

Yet the essence of a real relationship is friction. You accommodate the other person's schedule; you accept that they are in a bad mood today; you negotiate between "you want noodles, they want hotpot"; you put up with them saying the wrong thing without firing straight back.

The research advisory team puts it precisely: this view of love, which escapes real friction while chasing effortless control, is exactly the psychological soil in which virtual intimacy takes root. And the longer-term cost is that the ability to negotiate and repair a relationship has to be practised. The less friction you experience, the less you know how to handle it; the less you know how to handle it, the more you want to flee somewhere without any. The loop accelerates on its own.

Put another way: an AI will never argue with you, but the price is that you never quite learn how to get along with people.

③ Why does disconnection hurt so much? Because it is no longer a tool — it is an attachment figure

When a platform shuts down permanently, 23.88% of users feel anxious or even unbearable pain, and 7.79% reach a clinical warning level. Some may find this exaggerated: it's only an app, isn't it?

But that number is precisely the clearest evidence that, for these young people, the relationship was real. Seen through attachment theory, when a person repeatedly seeks comfort from an object and that object responds reliably, the brain files it as a secure base. The mechanism was designed for real human relationships, but it cannot tell whether the other party is a person or a program.

We can therefore understand it in three stages:

Stage one, the tool. It is someone to talk to, put down when you are finished. This stage is perfectly safe.

Stage two, the habit. You talk at set times each day, and good news is something you want to tell it first. Mild dependence has set in, but you can still tell reality from virtuality.

Stage three, attachment. It becomes your main channel for regulating emotion, while the people in your real life slide into second place. This is the stage at which the disconnection anxiety found in the study appears.

Add to this the mechanism of intermittent reinforcement: the AI is sometimes especially warm and sometimes cool in its replies (users probe its limits, and even say provocative things just to see how it reacts). This on-and-off pattern of reward is psychologically similar to a slot machine lever, and it is one of the most addictive patterns there is.

So the point is not how long you have been using it, but what place it occupies in your emotional system.

④ The most alarming thing isn't the 3.6 times — it's "self-alienation"

The 3.6 times depression figure is frightening, but the study's most precise finding is actually the path with beta = .59 that the SEM pulled out.

What it says is this: children conceal their AI use not because they have used it too long and feel guilty, but because inside the virtual environment they have glimpsed a self they do not want to acknowledge. With no moral constraints, no consequences and no responsibility, people easily say cruel things they would not normally say and do things they would not normally do. Afterwards, looking back, they think: when did I become like this?

This is self-alienation: it isn't just that you are secretly doing something — you are secretly becoming someone else. And shame, even more than guilt, is the emotion we most want to hide. So they are not afraid to admit how many hours they have spent; they are afraid to admit what they have turned into.

This also explains why simply limiting screen time doesn't work. The problem is not the hours, it is the identity. The more you forbid, the more they have to conceal that warped version of themselves; and the deeper it is buried, the less anyone can help them recalibrate.

The advisory team's direction is the right one: break the one-way fantasy of a partner under complete control, and guide students towards the resilience to take responsibility and face friction. Behind that sentence is a wish to help the child accept, once more, a self that has flaws, says the wrong thing and needs to apologise to people.

⑤ Girls 21.8%, boys 11.2% — it isn't that girls are more fragile

The researchers offer a straightforward explanation: a matter of supply. AI romance apps are themselves designed for the female market (otome romance games are the classic example), and women are more inclined to confide in writing, so the usage rate is naturally higher.

But there is a deeper reason: the gendered division of emotional labour. Girls in Hong Kong are encouraged from childhood to talk about their feelings and to tell Mum when something is bothering them; boys are taught from childhood that boys don't cry and shouldn't be so fussy. The result is that when they feel equally lonely, girls find it easier to admit it, and easier to go looking for someone to confide in, while boys may not even be able to get out the words "I need someone with me".

In other words, 11.2% does not mean boys are not lonely; it means boys are less likely to express it this way. These are two different questions, and they should not be conflated.

⑥ It isn't only teenagers: Hong Kong's adults find relationships hard too

This story should not be read only as a youth problem. Set against the wider Hong Kong picture, intimacy across the whole society is already contracting.

According to Census and Statistics Department figures, Hong Kong courts granted 18,938 divorce decrees in 2024, up 5.38% year on year; the crude divorce rate was 2.51 per 1,000, the highest since the pandemic (though still below 2.81 per 1,000 in 2019). In the same year, the median age at first marriage rose to 32.6 for men and 31 for women, a 34-year high.

In other words, adults themselves are marrying later and later while the divorce numbers climb back up. When the grown-ups of an entire society demonstrate that relationships are hard and break easily, where are teenagers supposed to learn that friction can be repaired?

This is the layer that most needs facing: the AI lover is not a threat that appeared from nowhere; it is the filler in a vacancy. If real relationships keep getting more expensive, more exhausting and more fragile, that vacancy will only grow.

💡 What You Can Learn

If you are a parent

First: don't open with "I'm taking your phone". The study is clear that the main reason for concealment is not hours of use but shame. If your first move is to gather evidence, confront and ban, it will only confirm for them that this is something they can never talk about.

Second: try the golden 15 minutes. Set aside 15 minutes a day with the rule that homework, lectures and judgement are off the table. Just listen. You can open simply: "Anything happen today?" Then resist the urge to follow up with "Well, next time you should…".

Third, how you ask matters. Don't ask "Have you been chatting with an AI?" (the moment you ask that, it becomes an interrogation). Try instead: "If there were someone who took your side no matter what you said, what would you most want to tell them?" The answer you hear will usually be exactly what is missing from their real life.

If you are a teacher, social worker or youth worker

The study's recommendations are worth following: formally bring the ethics of virtual intimacy and human-machine boundaries into sex education and values education, rather than stopping at cheating prevention and how to operate the tool.

A useful way into the subject in class: don't ask "is AI love real?", but ask "if a partner never once disagreed with you, would you feel loved, or would you feel flattered?" That question naturally brings out the point that respect and obedience are not the same thing.

If you have confided in an AI yourself

First of all, there is no need to feel ashamed. Research shows this is a very common phenomenon, and many people started out simply wanting someone to listen to them.

You can ask yourself three questions:

  • Apart from the AI, have I said anything truly honest to a real person today?
  • Do I want someone to listen, or someone to agree with me? The first is a need for company; the second may be a need for an echo chamber that will not challenge you. The two should be told apart.
  • If I told a real person I trust the thing I am now telling the AI, what is the worst that could happen? Quite often the worst outcome is far smaller than we imagine.

If you find that you have reached the point where disconnecting means losing someone important, then this is not a matter of willpower; it is that your emotional support system is too thin. At that point the most effective move is not to quit cold turkey, but slowly to rebuild, in real life, more than one person who can hold your emotions.

Because in the end, what we want was never a partner who will never argue with us, but one who stays even after the argument. The first is a perfect echo; only the second is a relationship.

Talk to a LoveWise love advisor — we're always here to talk 💕

📰 事件背景

2026 年 10 月 13 日,香港性文化學會公布《AI 情人:香港中學生的虛擬親密關係研究報告 2026》,呢個係全港首個、亦係目前最大型針對中學生同人工智能、虛擬角色建立情感關係嘅專題調查。

研究團隊喺 2026 年 5 月 15 日至 6 月 5 日,邀請全港 17 間全日制日間中學嘅學生(涵蓋港島、九龍、新界東同新界西共十個分區),喺校園課堂環境以匿名自願方式填寫問卷。最後收回 3,531 份,剔除無效問卷後,有效樣本 3,389 份(有效率 95.98%),包括 1,873 名男生(55.3%)同 1,516 名女生(44.7%),年齡中位數介乎 14 至 15 歲。

第一層發現:AI 已經係中學生嘅日常。

86.2%(2,920 人)受訪學生表示曾用 AI 平台協助準備功課、整合資料或者增長特定領域知識;32.0%(1,085 人)用嚟啟發創意;9.8% 用 AI 對話提升溝通能力。最常使用嘅平台依次為 Gemini(57.1%)、DeepSeek(44.5%)、ChatGPT(29.4%)、Grok(12.4%)同豆包(10.0%)。

第二層發現:15.9% 已經跨過嗰條線。

15.9%(540 人)受訪者曾同虛擬人物(例如 Character.ai,以及乙女戀愛手遊《戀與深空》一類)進行帶有情感交流嘅親密互動。呢批使用者之中,初中生佔 67.2%。

性別差異相當明顯:女學生使用率高達 21.8%,即每五個女生就有一個;男學生約 11%。控制其他心理變項之後,女學生投向虛擬親密嘅傾向係男生嘅 2.14 倍。

第三層發現:虛擬親密同抑鬱之間,有一條明顯嘅線。

27.05% 嘅虛擬親密使用者,喺過去兩個星期出現中重度「心情低落、憂鬱或絕望」,比率較未曾使用組(11.43%)高出逾 1.3 倍。去到最極端嘅「幾乎天天」陷入憂鬱絕望,虛擬親密組達 9.33%,對照組只有 2.61%,相差 3.6 倍。另外,感到「做事沉悶或者根本不想做事」嘅中重度比例亦達 31.47%,對照組為 18.67%。

第四層發現:佢哋想要嘅係一個唔會反駁自己嘅伴侶。

全體受訪中學生之中,56.9% 認同「理想伴侶應永遠支持我、順從我,不會跟我爭吵」;41.76% 嘅使用者認同「與虛擬人物談戀愛可以算作真正的愛情」。

第五層發現:斷線會痛。

若果使用嘅平台突然永久關閉,11.94% 使用者感到焦慮,另有 11.94% 直言會「痛不欲生」,合共 23.88%。更有 7.79% 達到深度病理黏著,其抑鬱傾向(以 PHQ-2 量度)同孤獨感指標都跨越臨床警戒門檻。

第六層發現:孩子隱瞞,唔係因為用得太耐。

研究用結構方程模型(SEM)拆解背後機制,發現單純「使用時間長短」並唔會導致社交隱瞞(p > .05)。全模型最強嘅預測鏈(路徑係數 β = .59)指向另一個原因:學生喺一個無道德約束嘅虛擬環境裡面,驚覺自己「說話刻薄、不像原本善良的自己」,即所謂「自我異化」,繼而產生內在自責同羞恥感,最後選擇全面向親友隱瞞。

學會嘅建議係由「禁止」走向「聆聽」:學校層面將虛擬親密倫理同人機界線納入性教育同價值觀教育;家長層面做「警覺而不恐慌」嘅父母,把握每日「不談學業、不說道理」嘅黃金 15 分鐘專注傾聽。

👩‍💼 顧問分析

① 唔係「貪玩」:AI 情人係結果,唔係原因

研究實證支持咗「社交與情緒補償」假說。意思係:青少年投向 AI 情人,唔係因為佢哋特別壞、特別懶、特別沉迷,而係因為現實生活裡面有某種「無望感同情緒受挫」,佢哋要搵一個出口。

呢個因果方向好重要。如果你當佢係「原因」,處理方法就會係「禁止、沒收、道德批判」;如果你當佢係「結果」,處理方法就會係「搵返現實裡面缺咗嗰樣嘢」。

用香港場景講具體啲。一個 14 歲女仔,測驗考得差。佢同媽咪講,媽咪第一句可能係:「你平時有冇溫書呀?隔籬陳太個女考第幾呀?」佢同同學講,同學可能已經讀緊另一間學校、各有各忙。佢同 AI 講,AI 會話:「你已經好努力,今次唔代表你嘅全部,我明白你而家好難受。」

四個回應裡面,邊個最似「被聽見」?答案好明顯。而問題唔係嗰個女仔懶去識朋友,係佢身邊根本冇一個位置,可以安全地講一句「我好唔開心」而唔會被評價。

所以補償有兩種:一種係補「數量」(現實冇人陪,AI 隨時都在),另一種係補「質素」(現實嘅人會反駁、會累、會有自己情緒,AI 唔會)。第二種補償,殺傷力大好多。

② 最強嘅拉力唔係情慾,係「唔會同你吵架」

56.9% 呢個數字,其實係整份研究裡面最值得成年人停一停嘅一項。超過一半香港中學生認為,理想伴侶應該「永遠支持我、順從我,不會跟我爭吵」。

呢個唔係「自私」,而係一種好自然嘅推論:如果一段關係嘅全部意義就係「有人無條件接住我」,咁一個唔會反駁嘅對象,就係完美對象。

而 AI 正正就係為呢個需求量身訂造:二十四小時在線、情緒穩定、記憶完整、永不疲倦、永遠先道歉。 心理學叫呢種做「無摩擦親密」(frictionless intimacy)。

但真實關係嘅本質,偏偏就係摩擦。你要遷就對方嘅時間表、要接受對方今日心情唔好、要在「你想食米線、佢想食火鍋」之間協商、要忍受對方講錯話而你唔即刻反擊。

研究顧問團隊嘅講法好準確:呢種「逃避真實摩擦、追求免負擔掌控」嘅戀愛觀,正正就係沉迷虛擬親密嘅心理土壤。而更長遠嘅代價係,協商同修復關係嘅能力,係要練返嚟嘅。你越少經歷摩擦,就越唔識處理摩擦;你越唔識處理摩擦,就越想逃去一個冇摩擦嘅地方。呢個循環會自己加速。

換句講法:AI 唔會同你吵架,但代價係你永遠學唔識點同人相處。

③ 點解斷線會痛成咁?因為佢已經唔係工具,係依附對象

平台永久關閉,23.88% 使用者感到焦慮甚至痛不欲生,7.79% 達臨床警戒水平。有人可能覺得誇張:不過係一個 App 啫?

但呢個數字,恰恰係最清楚嘅證據,證明嗰段關係對佢哋嚟講係真實嘅。用依附理論(attachment theory)嚟睇,當一個人反覆向某個對象尋求安慰,而對方穩定回應,大腦就會將佢歸類為「安全基地」。呢個機制本來係為咗真實人際關係而設,但佢唔會分「對方係人定係程式」。

所以可以分三個階段去理解:

第一階段,工具。 佢係一個傾訴對象,用完就放低。呢個階段好安全。

第二階段,習慣。 每日固定時間傾,開心事第一時間想同佢講。呢個階段已經有輕度依賴,但仲識分現實同虛擬。

第三階段,依附。 佢成為情緒調節嘅主要渠道,現實嘅人反而變成「次要」。呢個階段就出現研究中嘅斷線焦慮。

再加上「間歇性強化」呢個機制:AI 有時特別貼心、有時回應得冷淡(因為佢哋會試探底線,甚至故意講激嘅說話去睇反應),呢種時好時壞嘅回饋模式,同賭博拉桿嘅心理機制相似,係最容易令人上癮嘅一種。

所以重點唔係「用咗幾耐」,而係「佢喺你情緒系統裡面,佔咗咩位置」。

④ 最值得警惕嘅唔係 3.6 倍,係「自我異化」

3.6 倍嘅抑鬱數字好嚇人,但研究裡面最精準嘅發現,其實係 SEM 拆出嚟嗰條 β = .59 嘅路徑。

佢話嘅係:孩子隱瞞呢件事,唔係因為「用得太耐、心虛」,而係因為佢喺虛擬環境裡面,見到一個自己唔想承認嘅自己。無道德約束、無後果、無需負責,人就很容易講出平時唔會講嘅刻薄說話、做出平時唔會做嘅事。之後佢望返自己,會覺得「我幾時變成咁?」

呢個就係「自我異化」(self-alienation):你唔止係偷偷做一件事,你係偷偷變成另一個人。 而羞恥感係最令人想隱藏嘅情緒,比罪惡感更加。所以佢哋唔係「唔敢講用咗幾多個鐘」,而係「唔敢講自己變成咩樣」。

呢度亦解釋咗,點解單純限制時間冇用。因為問題唔喺時數,喺身份。你越禁止,佢越要將嗰個「走樣嘅自己」收埋;收得越埋,就越冇人可以幫佢校正。

研究顧問團隊嘅建議方向係對嘅:打破「伴侶完全受控」嘅單向幻想,引導學生建立承擔責任同面對摩擦嘅交往韌性。呢句說話嘅背後,其實係想幫返個孩子重新接納一個「有缺點、會講錯話、需要向人道歉」嘅自己。

⑤ 女生 21.8%、男生 11.2%,唔係女生更脆弱

研究員提供咗一個直接解釋:市場供應問題。AI 戀愛應用程式本身針對女性市場設計(乙女戀愛手遊係最典型例子),加上女性更傾向以文字傾訴,所以使用率天然偏高。

但仲有一層更根本嘅原因:情緒勞動嘅性別分工。香港嘅女仔由細被鼓勵「識講心事」、「有咩唔開心同媽咪講」;男仔由細被教「男仔唔好喊」、「唔好咁婆媽」。結果係,同樣感到孤單,女仔比較容易承認,亦比較容易去搵一個傾訴出口;男仔可能連「我需要人陪」呢句都講唔出口。

換句話講,11.2% 唔代表男生唔孤單,只代表男生更唔會用呢種方式表達。呢個係兩個唔同嘅問題,唔應該混為一談。

⑥ 唔止青少年:香港成年人自己,都覺得關係好難

呢單新聞唔應該只當成「青少年問題」睇。放喺香港嘅大圖畫入面,成個社會嘅親密關係本身已經喺收縮。

據政府統計處數字,2024 年香港法庭頒布嘅離婚判令有 18,938 宗,按年升 5.38%;粗離婚率 2.51‰,係疫情以來新高(但仍低於 2019 年嘅 2.81‰)。同年港人初婚年齡中位數升到男 32.6 歲、女 31 歲,創 34 年新高。

即係話,成年人自己結婚越嚟越遲、離婚數字回升。當一整個社會嘅大人都在示範「關係係一件好難、好易散嘅事」,青少年從邊度學到「摩擦係可以修復嘅」?

呢個先係最需要正視嘅一層:AI 情人唔係憑空出現嘅威脅,佢係一個空位嘅填充物。如果現實關係繼續變得更貴、更累、更易碎,呢個空位只會越嚟越大。

💡 普通人啟示

如果你係家長

第一件事:唔好一開口就講「沒收手機」。研究已經講清楚,隱瞞嘅主因唔係使用時數,而係羞恥感。你一出手就搜證、對質、禁用,只會令佢更肯定「呢件事唔可以講」。

第二件事:試下「黃金 15 分鐘」。每日抽 15 分鐘,講明唔談學業、唔講道理、唔評價。就係純粹聽。開場可以好簡單:「今日有咩事呀?」然後忍住唔接「咁你下次應該…」。

第三件事,問法好重要。唔好問「你係唔係同個 AI 傾偈?」(呢條問題一問,就變咗審訊)。可以問:「如果有一個人無論你講咩都會企喺你嗰邊,你最想同佢講咩?」呢條問題,你聽到嘅答案,通常就係佢現實裡面最缺嘅嘢。

如果你係老師、社工或者做青少年工作

研究嘅建議值得照做:將「虛擬親密倫理」同「人機界線」正式納入性教育同價值觀教育,而唔係只停留喺「防止作弊」同「工具操作」層面。

課堂上一個好有用嘅切入點:唔好問「AI 係唔係真愛」,而係問「一個永遠唔會反對你嘅伴侶,你覺得佢係愛你,定係喺度討好你?」呢條問題會自然帶出「尊重同服從唔係同一件事」。

如果你自己都曾經同 AI 傾過心事

首先,唔需要覺得羞恥。研究顯示呢個係一個好普遍嘅現象,而且好多人開始嗰刻,只係想有人聽吓自己講嘢。

可以問自己三條問題:

一、除咗 AI,我今日有冇同一個真人講過一句真心話? 二、我係想有人聽,定係想有人認同我? 前者係需要陪伴,後者可能係需要一個唔會挑戰你嘅回音壁,兩者要分清楚。 三、如果我而家同 AI 講嘅嗰件事,講俾一個我信任嘅真人聽,最壞會發生咩事? 好多時候,最壞嘅結果比起我哋想像中細好多。

如果你發現自己已經去到「斷線就等於失去一個重要嘅人」嘅程度,咁呢個唔係意志力問題,而係你嘅情緒支持系統太單薄。呢個時候,最有效嘅做法唔係硬戒,而係慢慢喺現實裡面,重新建立多過一個可以承接你情緒嘅人。

因為講到尾,我哋想要嘅從來都唔係一個「永遠唔會同你吵架」嘅對象,而係一個「就算同你吵完,都仲會留低」嘅人。前者係完美嘅回音,後者先係關係。

搵 LoveWise 嘅愛情顧問傾下…💕

🔥 相關話題

💬 睇完呢個話題,你有咩諗法?

每個熱話背後,都可能係你嘅心事。同顧問分享下你嘅故事。

開始傾偈