A more charitable interpretation is that the text that people thought would best train ChatGPT tended to be thoughtful and well-written posts like yours. Maybe you don’t write like ChatGPT, but ChatGPT writes like you.
As a non-AI Learning Model, I cannot conclude one way or the other with any certainty. What I can say is that ChatGPT responses tend to follow a similar pattern:
Consistent and clear responses: ChatGPT will often respond to prompts with very readable, well-formatted bulleted lists
Socratic reasoning: Items in those lists will have a logical structure from beginning to end
Finally, ChatGPT responses tend to end those lists with a summarizing statement that restates the previous ideas - that ChatGPT will often respond in lists, use a formal and logical writing style, and end with a concise summary of the previous statements.
“Don’t give me your mumbo-jumbo Mister Scientist - will the alarm go off or not?!”