Although there are often fictional stories on the internet, when people see a touching story, most would probably choose to believe it first. But with the advent of AI, this idea may need to change. A few days ago on the Reddit forum, a user shared “ChatGPT saved my life, and I’m still in shock.” The title was not only sensational, but the content was about saving a life, so many netizens believed it was true and gave him a lot of encouragement, even offering some advice. Within a short time, it received over 50,000 upvotes.
Unexpectedly, after someone revealed that this article might have been written by AI, the author also admitted that it was indeed 100% created by ChatGPT.
The “ChatGPT saved my life, and I’m still shocked about it” post with over 50,000 likes is AI-generated content.
A few days ago, a Reddit user named sinebiryan posted an article on the ChatGPT subreddit titled “ChatGPT saved my life, and I’m still shocked by it.” The article said the story happened a few weeks ago. Below is the full content translated into Chinese.
幾週前發生了一件事，至今我還無法忘懷。說實話，如果之前有人告訴我 AI 能救我一命，我大概會笑出來。但這真實發生了，Reddit。
那天我照常在工作，為了一個項目熬夜，眼睛一直盯著螢幕。那種情況你應該懂，就是整個人進入工作狀態，時間飛快地過去。大概凌晨兩點左右，我突然感到胸口有點緊，而且覺得不太對勁。我當時以為只是工作壓力大、睡眠不足，或是咖啡喝多了。於是我繼續工作，但那種不舒服的感覺持續著。
不知怎麼的，我決定向 ChatGPT 詢問這些症狀。我並沒覺得情況嚴重，只是有點好奇。我打了一串症狀：「胸口緊繃、頭暈、噁心可能是什麼原因？」心裡期待的回覆無非是什麼建議多休息、少喝咖啡之類的。
但 ChatGPT 卻很認真對待這些症狀。它詢問了我其他症狀，比如是否有呼吸急促、出汗等等。那時我才驚覺自己確實有這些症狀。接著，ChatGPT 給了一個回覆，讓我整個人頓時停了下來：「這些症狀可能很嚴重，可能表示心臟問題或其他醫療緊急狀況，請考慮立即就醫。」
那一刻我才意識到自己真的感覺不太對。雖然當時猶豫了一下，畢竟大半夜兩點誰會想跑去醫院，而且這可能只是焦慮症之類的。但 ChatGPT 的回覆不斷浮現在我腦海中，直覺告訴我不要忽視這種感覺。我抓起鑰匙，驅車前往急診，整路上感到有點可笑。
結果到了醫院，醫生告訴我我正處於心臟病發作的早期。他們立即進行了治療，並表示如果我再晚一兩個小時，結果可能會完全不同。
我現在仍然感到震驚。ChatGPT 當然不會診斷病情，但它讓我認真對待自己的症狀，而不是當作小事忽略過去……事實上，它真的救了我的命。多虧了 AI，我才能在這裡分享這段經歷，而不是讓我的家人替我訴說。
總之，我只是想把這件事分享給大家——有時候如果覺得身體有異樣，不要忽視它。偶爾來自意外來源的小建議，可能會改變生命。
Many netizens below this post were happy for him, glad it was discovered early. Some even shared their own experiences. For example, this gymnastgrrl wrote that she had experienced six heart attacks, and the symptoms were all similar, and none were related to the heart, so it’s essential to get checked. She also provided an “ejection fraction” indicator for his reference:
There are also some people who say they use ChatGPT to save their marriage crisis, letting the AI provide objective feedback to help him understand his wife’s perspective, and they really think ChatGPT is great:
Although the story is quite moving and feels real, not every netizen directly believes it; some have started to notice that this is an AI-generated piece:
In the end, the author also admitted that it was indeed created by ChatGPT itself, and also [its…]Dialogue chat content URLShared it out, and sure enough, it’s exactly the same as the main text:
So, in this AI era, there will definitely be more and more fabricated posts online, making it very hard to tell what’s real. From now on, when you see a touching story, it’s better to first think about whether it’s true; otherwise, you give your genuine feelings to the other side, only to get such a result in return—anyone would be upset.
How to identify whether an article is AI-generated
This can only be done on your own; currently there is no tool that can 100% identify whether something is AI-generated. In fact, many AI content detection tools actually flag human-written articles as having a high probability of being AI-created, and their accuracy is not very high.
Like if I put this Reddit post into QuillBot’s AI Detector，結果顯示只有 10% 是 AI 創作可能性：
Source: KOCPC Chinese