Using AI as a Tool for Writing Amidst Complexity

Medium.com recently asked its readers: We want your feedback: How can writers use AI to tell human stories?

My take is that artificial intelligence is probably mankind’s most complex creation since the internet. Hence, there is no simple way to explain the dynamic learning happening between humans, AI and nature.

The question then becomes: When are we telling a story and when are we describing reality? We can interpret the quality of AI’s responses and potential for facilitating skill practice while also acknowledging the dangers and risks that come along with it. Medium’s article is diving into: “How we are navigating the role AI does and doesn’t play in our mission to deepen understanding.” This is a good place to start because artificial intelligence may do an excellent job explaining yet it will never understand the human experience.

Still, the challenge will be in having hard discussions about hard problems. And when able, embracing complexity. For example, Medium’s post on this topic is relevant. But when it wrote: “The vast majority of AI-generated writing is bad and lazy slop, we can all agree.” I had to ask, “Why frame the context this way?”

“Bad”? “Lazy slop”? “We can all agree”?  C’mon Medium! Such simplification attempts to reduce the complexity of the context.

AI can write you a sonnet as if it was Christopher Marlowe, alive and well. In a matter of two seconds. Though writing may be subjective, the worst I’ve personally seen AI write in 2025, I would describe as: slightly above mediocre. Whatever description or hyperbole you want to use, the fact is the quality of AI will only continue improving and quickly.  Which is why, what is most concerning about AI in my opinion, is not whether it generates low quality writing (it doesn’t) or whether it’s useful to writers (it is) but addressing concerns on privacy protection. Especially, with biometrics and other forms of our data being sold and used for surveillance.

That being said, here is another insight from my own work. I always begin a project by writing my first draft or its first portion without AI’s critiques. This action maintains the integrity, direction and tone of the message. (Unless you’re conducting an experiment just for shits and giggles.) Then, I feed my 100% human-made draft to multiple AI for them to evaluate. (One fun initial prompt is: “Rate my first draft on a scale from 1-10 you cannot use 7”.) Afterward, I synthesize their feedback for reflection during the revising and editing process. Finally, using my own perspective and considering the purpose for the piece, I either accept, reject, or dismiss their considerations. Since, at the end of the day, good taste and emotional resonance is uniquely human. Plus, I always have the option to just not use AI at all. I can communicate in a way robots cannot, heart to heart. Like I did with this piece. (I published this text without sharing it with any LLMs.)

Of course, there is an infinite amount of ways to reach great writing. Yet, the tinkering a writer can now do with AI allows for different approaches and styles to emerge. Returning to complexity, AI equips us to uncover more, even if it’s just researching concepts during brainstorming. But that’s not learning in the truest sense. Learning is action and demonstrating skills. In this case, creating, composing, clarifying, expressing and delivering the truths of our vision or experience.

It’s not about what the author knows. It’s about how he lives, moves, and hones his craft. It’s like what Charlie Rose once said in an interview on the written word: “Don’t tell me you want to be a writer. Tell me you want to write!”