ChatGPT: A Brilliant Computer Brain With Outdated Data, A Broken Link Style, and No Clue Who the President Is…
“Copy and paste is the foundation of the web. Break that, and you break trust.”
The Promise of AI in News and Writing
Artificial intelligence has already reshaped how we work, research, and publish. For journalists, bloggers, lawyers, and students, ChatGPT can feel like a revolutionary tool, a digital assistant that can brainstorm, draft, and analyze at lightning speed. It’s not hype to say this is one of the most transformative technologies of our time.
But here’s the catch: the technology is smarter than the people managing it. The developers behind ChatGPT have made choices that betray a lack of understanding of how writers, publishers, and professionals actually use the internet. The result? A powerful tool hampered by broken basics.
Copy and Paste: A Broken Foundation
The most glaring flaw is also the most absurd. Copy and paste, the most basic feature of the internet, is broken inside ChatGPT. Instead of delivering clean, usable links, ChatGPT forces them into “prettified” cards or Markdown-style links. It looks nice in a chat window, but it’s a nightmare in practice:
Bulk copying links doesn’t work often you only get one link at a time.
Multiple links are mashed together in useless formatting.
Sometimes links redirect to a homepage instead of the actual source page.
For any professional writer, that’s not just inconvenient. It’s unacceptable. Credible reporting and SEO-driven publishing depend on fast, clean sourcing. When you can’t even copy your sources properly, the entire workflow breaks down.
Accuracy Failures That Undermine Trust
ChatGPT also struggles with factual accuracy in ways that should never happen in 2025. Even today, it often insists Joe Biden is president and refers to Donald Trump as “the former president,” regardless of the current reality. For casual users, maybe this seems like a minor glitch. For journalists, it’s a disaster. Accuracy isn’t optional it’s the bedrock of credibility. If the model isn’t laser up-to-date, it risks spreading misinformation with every generated story.
Fake Quotes: A Dealbreaker for Newsrooms
Worse still, ChatGPT has a tendency to fabricate quotes, even when explicitly told not to. This isn’t a small error it’s a credibility death sentence. Fake attributions poison journalism. They turn powerful reporting into sloppy misinformation and erode reader trust instantly. No newsroom, no serious publication, can risk an AI tool that slips in fake quotes. It needs a hard stop: if a quote isn’t real, the model should refuse to generate it. Period.
Developers Out of Touch With Writers
The common thread here is simple: the people building ChatGPT are not writers, editors, or journalists. They’re software developers, and their design choices prove it. For Example everyone on staff as a writer has more writing experience on ChatGPT than the people working on it do. It would behoove them to partner with a news team like ours ASAP.
They prioritize “delightful” formatting over professional usability.
They underestimate the necessity of clean copy-paste for writers.
They allow fabrications because they don’t live in industries where a single error can destroy credibility.
It feels like the AI itself is smarter than the human decisions that hobble it. Instead of unleashing its potential, the system is held back by misguided human design choices.
Fixes Are Simple, But Urgent
These problems aren’t unsolvable. In fact, they’re basic:
Fix copy and paste. Links must copy cleanly, in bulk, with no formatting tricks.
Stop “pretty” links. Deliver true blog-style hyperlinks or raw URLs, not gimmicks.
Stay current. The model must know who is president, what’s happening now, and avoid stale information.
End fake quotes. If it can’t be verified, don’t generate it.
The Verdict: Stop Designing for Toys, Start Designing for Pro Tools
ChatGPT could be the most indispensable tool in modern journalism and writing. But right now, it feels like a Ferrari with square wheels, full of potential, crippled by decisions made by people who don’t use it the way professionals do.
Every broken copy-paste, every outdated “fact,” every fabricated quote is a reminder that the brilliance of the AI is being dragged down by human error.
If OpenAI wants to win the long game, it needs to stop treating ChatGPT like a toy for casual users and start treating it like the professional tool it’s meant to be. Writers, journalists, and creators don’t need “delight.” They need accuracy, reliability, and functionality that works. Until then, ChatGPT will remain a genius in chains — brilliant, but held back by the very people who built it.
