Prompted 'almost human' content

[Graybeard]

I prompted almost human content labeled as "Maybe AI" this morning when I was doing a POC (proof of concept) for a demonstration post on Reddit.

I used some notes from 2013 about consumer reviews and their positive effect on sales as a concept.

prompt: Why are major companies are increasingly coming to realize consumer voices can wield enormous influence in shaping the opinions of other consumers?

The AI text as is, is attached as a *.txt file below.

ChatGPT3.5 scored 66% human. Either ChatGPT is learning from human input or the detector I'm using is getting dumber, I'm not sure which.

But it's an interesting conundrum, and every day we are seeing more and more posts that appear to be AI posers.

 

[T J Tutor]

every day we are seeing more and more posts that appear to be AI posers.

It's really out of hand. You and I do have our instinctual detectors from our years as businessmen and our years of experience in this industry having read likely millions of lines of text over the many decades. Trust your instincts brother!

ChatGPT is learning from human input

It has been doing that from the onset I think.

 

[AI Tool Directory]

We enjoy using +60% spartan in our prompt.