← Back to portfolio

Here's What Happened When I Used a Chatbot to Write a Blog Post

Published on

Estimate reading time: 1 1/2 minutes.

Since late last summer, there's been much chatter in about the ease of use and low cost of ChatGPT for writing content over the "cost" of human writers. Rather than see this rise of the chatbot as a frightening development, I began using ChatGPT for legitimate purposes other than writing and posting blog content. (I certainly don't use it write client content.)

But, I became increasingly curious about whether these generative AI writing applications that purport to have helped multiple top brand users produced good content. So, I did my own experiment using a ChatGPT-driven platform that I won't name.

I had the platform draft a story on a subject I know well—life insurance. From title to conclusion, the 'article,' Securing Your Children's Future: Why Moms Need Life Insurance More Than Ever," was it poorly written. There also was no research backing the assertions in the blog post. I always use research to support my business blog content, so that bothered me. 

But, when I used ZeroGPT to see if it detected AI-generated content, the results astounded me. Here's the output I got.

Some experts say that these AI detectors only suggest that content might be AI generated, since they're probably inaccurate most of time. But, I used AI to create the content, and I did no edits before running it through ZeroGPT, and these results are spot on. At least when I scanned this text for plagiarism using Grammarly, I found none. That means, to me, that AI didn't "steal" other people's content to write this.

Moral of the story? If you want to avoid having your content questioned as AI-generated, you'll have to write it yourself or hire a human to write—or at least substantially rewrite—the content. (Most GAI detectors attempt to tell you if content is 100% human generated, too. But, they're notoriously inaccurate.) Otherwise, expect poorly written content with no research support that may offend your readers

I can't see respected brands publishing such poorly written content that may damage their brand reputation. So, writers who offer data-backed, well-researched, expert-driven, engaging stories have no worries about losing their clients to AI right now. 

In fact, we might get offers to rewrite AI-generated content at our regular fees, and that's good for our business. However, I suggest writers review indemnity clauses in client contract related to rewriting AI-generated content. Legal experts say legal issues can arise for us from that work if we're not careful.

(c) 2023. Dahna M. Chandler for The Financial Communicator, Inc., a division of Thrive Media Collaborative, Inc. All rights reserved. This blog content may not be reproduced or reposted in whole or in part or used for AI applications without express written permission from the author. Reading time estimated by Read-O-Meter.

Image: Alex Knight on Pexels