- AI writing frequently uses the rhetorical device 'It's not X, it's Y' or negative parallelism, making it a telltale sign.
- This construction has significantly increased in corporate communications and AI-generated text, with chatbots like ChatGPT and Claude heavily relying on it.
- Experts theorize AI favors negative parallelism due to reinforcement learning from human reviewers and its role as a safe, predictable sentence structure.
- Fixing the issue is challenging due to AI models training on their own output, creating a feedback loop that perpetuates the device.
- While annoying, AI clichés help distinguish machine writing from human, though they risk making a once-effective rhetorical tool seem like a bot trademark.