
ChatGPT, crisis of em dash, and please leave our beloved punctuation alone
Yes, it is all because of ChatGPT and that crazy beast called popular taste. Many among you dear readers, many among you, will read the first two paragraphs of this piece and immediately think that it has been written by ChatGPT. Because somewhere you have read, or someone has told you, or maybe you have seen some X thread called How To Identify ChatGPT Writing, explaining that one tell-tale sign of AI writing is em dash. It is not without logic. ChatGPT and other AI tools do use em dash liberally. But what is not right is that just because an AI tool loves a punctuation mark, we are being forced to discard it.There is a reason. ChatGPT loves em dash because writers love it. Em dash is not a punctuation mark like semicolon, a reviled and pretentious sign that is neither here nor there. Semicolon deserves its ignominy. Samuel Beckett hated it so much that in his novel Watt, he ended up writing, 'how hideous is the semicolon.' And Kurt Vonnegut famously called it 'transvestite hermaphrodites' in an era when politically incorrect things could be said without a lynch-mob forming outside one's house. Vonnegut was clear: no semicolons. 'Here is a lesson in creative writing. First rule: Do not use semicolons. All they do is show you've been to college,' he once said.advertisementYes, semicolon deserves all the hate. But not em dash. Because em dash, along with comma, represents how we think. And how we think is how we write.In our regular conversations, such as when everyone is sharing gossip during dinner, we talk linearly. This means one word leads to another, one thought leads to the next, and there is a thread running through them. But writing is different. Writing is not speaking. Instead, writing is a far truer reflection of thinking. And when it comes to thoughts, it is easy to notice that our thoughts are not linear. Instead, thoughts are haphazard, and every time you think about how sweet mangoes are, you are also probably thinking that similarly sweet are melons.Writers love em dash because it can be used in tens of different ways to organise haphazard thoughts. It can be used to branch out to a new tangent — and how lovely that is, like wishfully thinking of sun on a rainy day — while staying the course. It can be used to connect two bits in a statement — join them together so they seem whole. It can be used to put emphasis on something — and think about this deeply — because it is an important point. It can be used to end a sentence with some force — stop. Its versatile nature makes em dash a favourite of writers.advertisementThis is also why ChatGPT — and other AI tools — learnt to use it. Various AI models have been trained in high-quality writing. Just a few weeks ago, we got reports that Anthropic bought millions of books and scanned them to train its Claude AI. The case is similar for ChatGPT and Gemini, or DeepSeek. You can say that AI has learnt writing from people like Hemingway and Joyse and Nabokov, the writers who used to love em dash.Given that AI has been trained more on writing, and less on speech, it uses punctuation that may seem out of place in 2025. We live in a time of podcasts and short videos and text messaging, a time when reading and writing are considered pretentious activities, when vocab of an average person is considered outsized if it has the word 'delve' in it. In this time, something like em dash may come across as a mark of artificiality even when it is not.advertisementPunctuation is a tricky subject. Rules exist, and more so in India, where we place our Wren & Martin on the same shelves where we keep our holy books, but in reality, these rules are extremely flexible. Their only task is to bring clarity and meaning to words. It doesn't matter in what manner they do it. This is the reason why Nobel Prize winner Toni Morrison struggled with her editor on commas. Her idea was that a comma is not for grammar, a thought that apparently did not sit well with her editor. 'He does not understand that commas are for pauses and breath. He thinks commas are for grammatical things. We have come to an understanding, but it is still a fight,' Morrison once famously said. It is also the reason why Gertrude Stein, who too disliked commas, wrote sentences such as 'Rose is a rose is a rose is a rose.'Writing is this complex, beautiful, shape-shifting thing. It slithers and simmers and, in the hands of good writers, it rarely follows, ahem, a script. But as AI takes over, there is a great rush to create new rules that can distinguish between AI writing and the words that humans have put together. The task is futile. Good writing is going to be good writing. And the marker for that goodness needs to be decided by merit and taste, and not on the basis of how many times em dash and the word 'delve' have been used in it.(Javed Anwer is Technology Editor, India Today Group Digital. Latent Space is a weekly column on tech, world, and everything in between. The name comes from the science of AI and to reflect it, Latent Space functions in the same way: by simplifying the world of tech and giving it a context)- Ends(Views expressed in this opinion piece are those of the author)Trending Reel

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


NDTV
an hour ago
- NDTV
WhatsApp Launches New Safety Tool, 6.8 Million Scam Centre Accounts Banned
New Delhi: WhatsApp has launched a new 'safety overview', which will alert a user when someone not in their contact list adds them to a new WhatsApp group that they may not recognise, as Meta's messaging platform intensified its crackdown on scams and fraud. The safety overview will contain key information about the said group and tips to stay safe, it said. "From there, you can exit the group without ever having to look at the chat. And if you think you might recognise the group after seeing the safety overview, you can choose to see the chat for more context," according to WhatsApp. And notifications from the group will be silenced until the user marks that they do wish to stay. Meta's messaging platform further said it is exploring ways to caution users when they start a chat with someone not in their contacts by showing more context about the person they are messaging. This, it hopes, will enable informed decisions. WhatsApp said it also takes down attempts by criminal scam centres, often fuelled by forced labour and operated by organised crime, primarily in South East Asia. In the first six months this year, as part of its efforts to protect people from scams, WhatsApp and Meta's security teams detected and banned over 6.8 million accounts linked to scam centres. "Based on our investigative insights into the latest enforcement efforts, we proactively detected and took down accounts before scam centres were able to operationalise them," it said. Citing an instance of such a crackdown, it said recently, WhatsApp, Meta, and OpenAI disrupted scamsters' efforts that had links to a criminal scam centre in Cambodia. These attempts ranged from offering payments for fake likes to enlisting others into a rent-a-scooter pyramid scheme or luring people to invest in cryptocurrency. "As OpenAI reported, the scammers used ChatGPT to generate the initial text message containing a link to a WhatsApp chat, and then quickly directed the target to Telegram, where they were assigned a task of liking videos on TikTok. The scammers attempted to build trust in their scheme by sharing how much the target has already 'earned' in theory, before asking them to deposit money into a crypto account as the next task," it said.


Deccan Herald
2 hours ago
- Deccan Herald
OpenAI eyes $500 bn valuation in potential employee share sale
ChatGPT maker OpenAI is in early-stage discussions about a stock sale that would allow employees to cash out and could value the company at about $500 billion, a source familiar with the matter said. That would represent an eye-popping bump-up from its current valuation of $300 billion, with the sale underscoring both OpenAI's rapid gains in users and revenue as well as the intense competition among artificial intelligence firms to secure talented workers. The transaction, which would come before a potential IPO, would allow current and former employees to sell several billion dollars worth of shares, said the source, who requested anonymity because the talks are private. Bolstered by its flagship product ChatGPT, OpenAI doubled its revenue in the first seven months of the year, reaching an annualised run rate of $12 billion, and is on track to reach $20 billion by year-end, the source added. Microsoft-backed OpenAI has about 700 million weekly active users for its ChatGPT products, a surge from about 400 million in February. The share sale talks come on the heels of OpenAI's primary funding round announced earlier this year, which aims to raise $40 billion, led by Japan's SoftBank Group. SoftBank has until the end of the year to fund its $22.5 billion portion of the round, but the remainder has been subscribed at a valuation of $300 billion, the source said. Microsoft-backed OpenAI has about 700 million weekly active users for its ChatGPT products, a surge from about 400 million in February.


Hans India
2 hours ago
- Hans India
Bridging Policing and AI: Nikeelu Gunda Trains Officers at PTC Medchal
In a progressive step toward integrating technology with frontline law enforcement, AI and Digital Strategy Expert Nikeelu Gunda conducted a special training session for police officers at the Police Training Centre (PTC), Medchal. The refresher course focused on the practical use of Artificial Intelligence and digital tools in everyday policing, cybercrime investigations, and responsible social media usage. During the session, officers were introduced to a range of real-world applications such as phishing scam detection, deepfake identification, mobile data analysis, voice recognition, digital evidence handling, and dark web monitoring. Nikeelu demonstrated how AI-powered tools like ChatGPT, Grabify, VirusTotal, and PimEyes can assist officers in detecting cyber threats and analyzing digital clues in real time. 'Technology has become the new beat for law enforcement,' said Nikeelu Gunda. 'Empowering officers with digital tools ensures they are equipped to tackle both on-ground and online crimes effectively.' The training was conducted under the leadership of PTC Principal P. Madhukar Swamy, who emphasized the importance of future-ready policing. 'As crimes become more digital, our police force must become more dynamic. This training is a vital step in preparing our officers for a data-driven world,' he stated. The event was successfully supported by Laxman DSP, along with Inspectors Kiran, Ravi, and N. Chandrasekhar, whose coordination helped make the session a success. Participating officers shared positive feedback, noting that they gained confidence in using AI to support investigations and handle digital complaints more effectively. The program marked another significant stride toward building a tech-savvy, cyber-aware police force in Telangana.