The Danger of Bullshit: AI Checkers are Nonsense

Problem with AI Checkers

ChatGPT is a useful tool for generating ideas, but it's usage is often misrepresented. Some AI checkers, such as Quillbot, claim to be able to detect whether an article was written by AI or not. However, these checkers can be flawed and often penalize well-structured and grammatically correct text.

False Positives

The article author had an incident where their work was flagged as 79% AI, but upon re-checking, the same work came back as 84% AI with no changes made to the text. This highlights the unreliability of these AI checkers.

Red Flags of Unreliability

• The checker flagging a paragraph as AI one minute and then fine the next minute.
• The checker flagging a paragraph as AI when no changes were made.
• The checker not flagging plagiarism as AI.

What to Do if Accused of AI Generated Content

Track your writing by using Google Docs, allowing you to view previous versions of the document. Add language to your contract stating that your work will not be detained by AI checkers. Educate your clients on the unreliability of these checkers.

Misuse of AI Checkers

• Removal of transition words and cohesive paragraphs.
• Reduction of headlines to unhelpful titles.
• Rewording to avoid exact phrases used in the writing industry.

Misconceptions about AI Checkers

• They are not unable to quell true plagiarism, only plagiarism in the form of AI.
• They have stood the test against well-researched plagiarism, not one-time plagiarism.
• It is not viable to attempt to locally block AI checkers, as without internet, you wouldn’t be able to express this.