I am a Highly Sensitive Dog when it comes to sniffing out ChatGPT3. This time I caught the well-known online magazine Psychology Today.
My detector usually triggers based on keywords, and on specific sentence/paragraph constructions. Plus, it reacts a certain slightly exaggerated tone.
When I ask the AI Claude about the differences between its own tone and the tone of ChatGPT, a part of the reply is this:
I may tend to be somewhat more direct and concise, especially for straightforward queries, while still providing thorough responses to more open-ended questions. ChatGPT might lean a bit more verbose.
So here are some favorite verbosity fillers of ChatGPT by which I recognise it everytime, if not run through some humanizing or cloaking AI.
To embark
To cultivate
Profound
Intricate
To transcend
Ultimately
(Rich) tapestry
Journey
Perversive use of "we"
Shocked by Psychology Today
Today I was reading an article about Artificial Intelligence in Psychology Today when my alarm bell rang.
Here is the text of a paragraph, in a text about AI and their potentially hypnagogic states.
Yet as powerful as this hypnagogic state may be, it's still vastly unexplored as a tool for creative enhancement or even discovery. And as we continue to explore the fascinating intersection of human and machine creativity, the hypnagogic state serves as a reminder of the incredible creative potential that lies within our own minds. By cultivating an appreciation for this unique state of consciousness and learning from the novel associations generated by AI, we may unlock new pathways to innovation and push the boundaries of what's possible in both human and artificial cognition.
It had allegedly been reviewed by a human. In the alleged review process, no mention is made of AI.
However, when I checked, Gptzero confirmed that the article has exactly zero (0%) human written content.
The author has an AI written biography
Ok, so, I look up the author, John Nosta. And what do I find in his bio? Two of my favourites:
Rich tapestry
To transcend
John Nosta's career paints a rich tapestry of intellectual curiosity, disruptive thinking, and innovative leadership. His important influence on technology, medicine, and creative strategy continues to shape and guide the future, placing him at the forefront of global thought and innovation. His multifaceted career encapsulates a tireless pursuit of knowledge and creativity that transcends traditional boundaries, establishing him as a true luminary in his field.
So let´s check with gptzero. Result: this passage is 100% AI-generated.
Conclusion: cheating by Psychology Today. John Nosta probably employs an intern that prompts ChatGPT to write its texts
Cheating? Yes. Psychology Today may be largely written by AI , under the name of authors whose bio is written by AI and who probably never see the text.
All the while pretending there is a thorough fact-checking and editorial review process.
I won´t read Psychology Today again, most likely, before they flag 100% AI as such.
Comments