Rethinking Analytics

I have a friend who publishes a blog. When we spoke last week, she commented that she no longer pays much attention to the statistics because she doesn’t think they are reliable or valid.

Her reasoning appeared quite sound. Before AI became a wide spread thing, her stats were down in the 30 to 200 views a day but since AI has expanded across the internet, she noted her views are usually between 600 and 3000 a day.

Think about it. Everywhere you look, people are offering courses on how to use AI to write books, articles, papers, and even create pictures and coloring pages. I know a school teacher who used AI to write articles for his newsletter so he didn’t have to do the work.

I know education is struggling with what to do with AI written papers in English and Social Studies among other subjects. I don’t know if the programs designed to find plagiarism work on AI composed papers. I don’t even know if you publish a book written entirely by AI, if it can truly considered the authors work or if AI is seen as a ghost writer.

Back to my friend and her blog. She feels that the stats are no longer a proper representation of the number of views since the chances most of the views are actually AI scanning the internet for material to write something.

If most of the views are indeed due to AI scanning the internet, how do you know the number of real people visiting your blog. I don’t know if analytics can determine if a view is AI or a real person as it seems the technology to determine things is always behind the creation of the item. So I see any software that can tell if something is written by AI or viewed by AI being developed long after AI becomes a valid tool.

This topic is quite intriguing and I wonder if they will develop analytics that can differentiate between AI and real viewers. I expect eventually it will happen. Let me know what you think, I’d love to hear. Have a good day.