Users have long complained about ChatGPT relying too heavily on em dashes, and OpenAI says it has now introduced a fix. The company has confirmed that the latest model, GPT 5.1, allows people to set custom preferences that tell the chatbot to avoid using the punctuation mark that became a familiar part of its writing style.
CEO Sam Altman said on a social platform that the update at last makes the assistant follow this instruction. He posted, “If you tell ChatGPT not to use em-dashes in your custom instructions, it finally does what it’s supposed to do!” and described it as a “small-but-happy win.”
The update matters to users who felt that the unusual frequency of em dashes revealed when a text was generated by artificial intelligence. Some writers even stopped using the punctuation in their own work to prevent doubts about authenticity. The change has also sparked renewed debate about the pace of progress toward artificial general intelligence. One user replied to Altman saying, “The fact that it’s been 3 years since ChatGPT first launched, and you’ve only just now managed to make it obey this simple requirement, says a lot about how little control you have over it.”
There is still no clear explanation for why language models tend to rely on em dashes. Researchers believe the punctuation appears very often in professional and older published writing that forms part of the model training data. It may also have been rewarded during reinforcement training if human reviewers viewed the punctuation as polished or refined.
The deeper challenge is that language models do not follow rules the way traditional software does. They predict words based on probability, so telling a model to avoid certain punctuation only makes it less likely, not impossible. Early reports from users show that the fix is helpful but not perfect, and occasional em dashes may still appear despite instructions.
The issue highlights a broader question. If it takes several years to control a simple punctuation preference, the path to achieving human level intelligence may be more complex than expected.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream formerly known as CIO News LinkedIn Account | The Mainstream formerly known as CIO News Facebook | The Mainstream formerly known as CIO News Youtube | The Mainstream formerly known as CIO News Twitter
About us:
The Mainstream formerly known as CIO News is a premier platform dedicated to delivering latest news, updates, and insights from the tech industry. With its strong foundation of intellectual property and thought leadership, the platform is well-positioned to stay ahead of the curve and lead conversations about how technology shapes our world. From its early days as CIO News to its rebranding as The Mainstream on November 28, 2024, it has been expanding its global reach, targeting key markets in the Middle East & Africa, ASEAN, the USA, and the UK. The Mainstream is a vision to put technology at the center of every conversation, inspiring professionals and organizations to embrace the future of tech.
