I do a lot of short writing “for an audience of one.”
Personal words of congratulations or encouragement to individuals who have posted on support lists.
I’ve been doing this for decades.
My most recent writing was 27 lines and took 45 minutes of careful thought.
Yesterday, I got an odd response:
Sounds like chatgpt 🤔😅
It bothered me a lot.
Today, it still bothers me.
It still feels odd.
Was my writing that poor?
Have my efforts to provide comfort missed its mark by that much?
Are my motives suspect?
I am told that on various social media platforms (and even here on LinkedIn), such questions are not uncommon, but I had not seen one before.
I have often said that the problem was not AI-generated content but rather content that was generated and targeted so badly.
But of course, there is another problem. The problem of work being wrongly believed to be AI-generated. I see it in classrooms. Damaging innocent students, wrongly accused. Disproportionately neurodivergent children.
I hadn’t thought about the false positive problem in the rest of the world.
I am embarrassed to have been fully oblivious to an important ethical and societal issue until I became a victim.
And I am embarrassed that my writing was that bad.
But most of all, I am sad.
#aiethics #ai #chatbot #chatgpt #falselyaccused #neurodivergent #kindwordsproject
