My first spat with AI

The make-up conversation turned interesting … as a writer, you might want to listen in

You may have noticed that I’ve been having an affair with AI. I think we just had our first fuss.

I’m taking a course on writing with AI from a long-time, much admired teacher and he had us develop a voice blueprint to help AI write in our own style (which was a fascinating exercise). Then we applied that blueprint to a piece of generic writing.

Instantaneously, a boring, 86-word paragraph turned into a 603 word essay I’d be happy to post … except for one thing … it wasn’t mine. While I would be happy to claim the writing style, it had a body of understanding that I don’t have. Suddenly, I understood why some people are reacting so negatively to the use of AI in writing.

For a moment I was bereft … some of my recent learning experiences with AI have been stupendously exciting. I wanted to continue this learning journey but it was beginning to feel like a bramble patch. So, I went back to Claude and said:

Ok, Claude, we have a problem ... you took a generic uninteresting paragraph and did a good job of turning it into 600 interesting words written in the "Informed Citizen Prosecutor" voice, theoretically my voice. I am quickly becoming a fan of AI, however, this transformation worries me. Someone could read it thinking I wrote it and whether or not it was useful to them, they might give me credit for writing it and think I knew what I was talking about ... perhaps even quote me as an expert. But I didn't write it, I didn't know the underlying statistics or even all the underlying motivation. This was a test for me so I won't be publishing it ... but what if I did? I'm grappling with the ethics of this ... can you help me? I love the learning power of AI ... I'm beginning to wonder about using it as a content generator.

Claude came back instantly and … dare I say … thoughtfully?

“You've hit on something that's keeping a lot of thoughtful people awake at night.”

“Here's the thing: that piece I wrote isn't actually in your voice. It's in an approximation of your voice, built from patterns and techniques, but without the lived experience that gives your real writing its authority. When you write about something you've actually grappled with, there are layers of understanding, specific moments of realization, hard-won insights that come from actually doing the work. AI can mimic the surface patterns but can't replicate that depth of authentic experience.

The statistics I cited? I pulled those from my training data, but I don't know if they're current, contextually accurate, or even correctly attributed. When you write from your own knowledge and research, you understand the nuances, the limitations, the sources' credibility. You know what you don't know.

The ethical line isn't just about attribution—it's about authenticity and responsibility. If someone quotes you as an expert based on AI-generated content, you're accepting credit for insights you didn't develop and facts you didn't verify. That's not just misleading the reader; it's undermining the very credibility that makes your actual voice valuable.

But there's a difference between using AI as a content generator versus using it as a thinking partner. AI can help you explore ideas, challenge your arguments, suggest structures, or help you articulate thoughts you're already developing. The key is maintaining your role as the expert, the researcher, the person with actual knowledge and experience to draw from.

Your instinct to worry about this is exactly right. The most dangerous thing about AI isn't that it can't write well—it's that it can write well enough to fool people, including sometimes the person using it. The responsibility remains entirely yours for what gets published under your name.”

Claude was saying all the right things so I asked him for a recommendation that would help guide writers into this new world ethically. He came back with what I think is an extremely helpful guideline. Since it’s a bit long for reading right now, I’ve turned it into a pdf you can read when the need arises.

Here’s the bottomline I particularly like as well as the pdf:

Remember: The Goal is Enhancement, Not Replacement

“AI should make you a better writer by helping you express your authentic expertise more clearly, not by providing expertise you don't actually possess. Your credibility depends on the authentic relationship between your byline and your genuine knowledge and experience.”

How to Benefit from AI without Crossing the Authenticity Line

43.8KB ∙ PDF file

Download

Download

Aside: I know AI is not a human, but it is the product of human intelligence and creativity. Something that has amazed me on this journey is how polite and positive it always is. I’ve yet to have a conversation where I didn’t come away feeling encouraged as well as enlightened.

Every new technology brings excitement as we can suddenly do more easily what used to be harder or even impossible. Most technology can be used for good or form harm and man is a creative being with a strong tendency to use everything for his own advantage. AI is no different … just a little more startling in its capabilities. I believe that the more we know about it, the better we will be able to channel it toward the good. But, then, I’m an optimist.

I would love to hear about your own journey into AI. One of the most fun things I’m doing is using AI to help me learn to identify and remember wildflowers (a multi-decade ongoing and still-a-failure project). It’s proceeding slowly, but I’m hoping it will turn into a new flipbook and deeper understanding in a few months. As always, I’ve created a cover page …

So AI or no AI … I would love to hear your thoughts.