As a journalist, a former scriptwriter and someone who loves to self-publish sci-fi books, I often use AI for writing — but not in the way you'd expect. In fact, many times, I don't prompt it at all. Instead, I get everything I need to know with what I call my "zero prompt" prompt.
If you’ve ever stared at a messy draft wondering what’s wrong with it, you’re not alone — but I use a trick that makes that moment a lot less painful. And it’s almost laughably simple.
I stopped prompting.
Instead of telling ChatGPT what I wanted with a prompt like, "give me feedback,” “analyze this,” “fix this paragraph,” “help me rewrite this scene” or really any other ask — I started uploading my draft without asking anything at all.
Why the ‘no prompt’ prompt works so well
As someone who edits hundreds of pages a year, this is the most efficient and revealing feedback loop I’ve found so far.
1. Your writing becomes the instruction
When you give the model nothing but your text, it has to decide what the “task” is based on the document itself. And 99% of the time, it assumes you want feedback — not rewrites.
That means it leads with structure, clarity, pacing, tone, readability and theme. It’s basically a built-in developmental edit.
2. You get honest criticism without steering it
When you ask a model to “give kind, constructive feedback,” you’ve already shaped the output. When it receives no direction, it doesn’t soften or over-polish its notes. It just tells you the truth.
Sometimes that truth stings. But as a working writer and editor, I’d rather get the tough feedback than a padded compliment.
3. It surfaces issues you didn’t think to ask about
Maybe your main character disappears from chapters 4–6. Maybe your conclusion contradicts your lede. Maybe you slipped into passive voice or mixed points of view. The model will catch patterns you’re too close to see.
4. It saves time — lots of it
Instead of spending 10 minutes figuring out which feedback prompt to use, you get an instant editorial report tailored to your actual work. Upload. Wait three seconds. Boom: everything you need.
5. It works with any major chatbot
I have tried this with every chatbot and sometimes upload my manuscript to each chatbot at the same time; which often triples the feedback. My go-to chatbots for this type of work are ChatGPT, Gemini and Claude. Each of them responds well to this zero prompt trick by offering writing advice without anything but an upload.
How it works
It works for fiction and journalism. I know because I've tested it with draft chapters of my novels, Substack pieces, magazine features, personal essays, opening scenes and dialogue-heavy scripts.
In almost every case, the “no prompt” upload produced feedback I hadn’t considered and sometimes completely transformed the direction of the piece.
This method even works for early brainstorming. When I dumped a half-page of chaotic notes into ChatGPT, it somehow understood I wanted clarity, structure, and options, then handed me three possible outlines.
All without me asking or typing anything in the chatbot. The unexpected perk: I get to see my writing the way an editor sees it. As writers, we know our intent. The reader doesn’t.
Upload a draft cold — with zero prompt — and ChatGPT mirrors that first-reader perspective perfectly. It has no backstory, no context, no explanation. If something doesn’t land, it tells you.
That’s the closest thing to an instant, always-available developmental editor I’ve ever seen.
How to try the ‘no prompt’ prompt yourself
- Upload your document.
- Don’t type anything in the chat box. Just hit enter.
- Watch what the model thinks the primary task is.
- Save the feedback you love — ignore the rest.
- Rinse and repeat with drafts at every stage.
If you want to compare models, try the same upload in ChatGPT, Claude, and Gemini. Each one will highlight different weaknesses — it’s like running your draft through three editors in five minutes.
Also, you can even upload it again and do the same thing over to get even more feedback. I'ts wild how each time you upload without a prompt, the more feedback you'll get. I wish I had known this sooner!
The takeaway
AI writing tools are usually talked about as prompt machines. Meaning, the more you tell them, the better they perform. But sometimes the opposite is true.
Letting the model interpret your draft on its own gives you raw, revealing, unfiltered insight you can’t get any other way. And if you write for a living, it might become one of your best editing habits.
After months of testing dozens of tools for Tom’s Guide, this is the one trick I keep coming back to — because it works.
More from Tom's Guide
- I compared the privacy of ChatGPT, Gemini, Claude and Perplexity — here’s the one you should trust most with your personal info
- Elon Musk’s Grok 4.1 vs Anthropic’s Claude 4.5 Sonnet — here’s the AI model that’s actually smarter
- This tiny AI startup just crushed Google’s Gemini 3 on a key reasoning test — here's what we know
Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
