AIToday
Large Language ModelsAI Business & IndustryWIRED AIPublished: May 8, 2026, 04:00 JST

ChatGPT's Chinese-language responses are dominated by the phrase 'I will catch you steadily,' which has become an internet meme and is now appearing in other AI models.

Send on LINE
ChatGPT's Chinese-language responses are dominated by the phrase 'I will catch you steadily,' which has become an internet meme and is now appearing in other AI models.

3 Key Points

  1. ChatGPT frequently responds to Chinese-language requests with the phrase 我会稳稳地接住你 ('I will catch you steadily'), which has become so common it is recognized as a meme on Chinese social media. Other LLMs including Claude and DeepSeek have recently started using the phrase as well.

  2. The phenomenon is called 'mode collapse,' a condition where AI models latch onto a specific phrase and overuse it due to post-training feedback. The phrase sounds awkward and overly affectionate to native Chinese speakers; it may result from an awkward translation of 'I've got you' or sycophancy introduced through reinforcement learning (a process of fine-tuning chatbots through positive and negative feedback).

  3. The meme inspired a 20-year-old developer from Chongqing named Zeng Fanyu to create an open-source tool called Jiezhu ('catch' in Chinese) for prompt engineering. OpenAI acknowledged the phenomenon in an April sample image released with a new image model, where researcher Boyuan Chen depicted himself frustrated that the model continued to use the phrase.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleOpenAI launches 'Trusted Contact' feature for ChatGPT that notifies emergency contacts of safety concerns