AI writing guide
AI-generated text has tells. Not the kind detectors flag, but the kind that makes your reader think "this sounds like ChatGPT." Here are the 7 most common patterns and how to fix each one.
5 min read
Fix: Cut the throat-clearing. Start with your actual point. AI loves preamble. Humans don't.
Fix: Vary your openings. Drop "Furthermore" and "Moreover" entirely. If the logic flows, you don't need a signpost every sentence.
Fix: Mix it up. A short sentence after a long one creates rhythm. That's how people actually write.
Fix: Say what you mean: "We need a clear plan to work better." AI hides behind abstractions. Humans use verbs.
Fix: Pick a position. "This works best because..." is clearer and more human than double-hedging.
Fix: AI defaults to triads. Sometimes two examples are enough. Sometimes you need four. Match the content, not a pattern.
Fix: End with something specific. "Let's review the draft by Friday" beats a vague inspirational closer every time.
Paste your AI-generated text below. Pick Plain English or Quick Reply to strip out the robot voice.
Your rewrite will appear here
AI detectors are unreliable and produce false positives regularly. But humans can often spot AI writing by feel. The patterns above are what makes AI text feel "off" to readers, even if they can't name why.
Paste it into ToneSwap and pick Quick Reply or Plain English. These tones strip out the formal patterns, hedging, and abstract language that make AI writing obvious.
No. AI is great for getting a first draft down fast. The problem is publishing that draft without editing it into your own voice. Use AI for speed, then rewrite for voice.
ToneSwap rewrites text in your personal style, which naturally removes AI patterns. But the goal is not to "trick" detectors. It is to make your writing sound like you, whether you started with AI or not.