Moving Beyond the AI Witch Hunt
A Non-Writer’s Perspective on Medium’s Future
In 1991 I was deeply engaged building “Expert Systems,” I can recall the panic, how these Frankenstein monsters were going to replace CEOs and much of the rank and file of the likes of Pepsi, International Paper, and John Hancock, all of which I developed AI to solve some problem. In doing this work I found one simple, very successful maxim: What Before How.
So I have to ask, WHAT is the concern with AI or more accurately ChatGCP like programs contributing to creativity on Medium?
The False Binary is Hurting Real Writers
The current discourse creates an artificial divide between “pure human” and “AI-generated” content that doesn’t reflect how most of us actually work. I’m 70+, dealing with dyslexia and ADD, and AI tools have become essential writing partners that help me get my ideas across clearly.
My thoughts, experiences, and insights remain entirely my own; AI simply helps me organize and articulate them more effectively. I am certain there are others not as fortunate as me who suffer from a greater degree of impairments (PSTD, ASD, BiPolar Disorder, Panic, Obsessive Compulsive and Anxiety Disorders) and they have very human stories to tell.
This isn’t about replacing human creativity; it’s about removing barriers that prevent people from participating in written discourse. When we frame AI assistance as somehow “dishonest,” we’re potentially silencing voices that these tools help amplify.
Medium Isn’t the Paris Review
I was a Medium reader for more than a year before I wrote my first article in 2018: I did not become a paying member until 2020. I have found Medium to be a valuable platform where people share ideas, experiences, and knowledge, not a literary magazine seeking the next Hemingway. The “AI slop detector” crowd is missing the point entirely. A well-researched, thoughtfully structured piece that used AI for research or editing assistance can be infinitely more valuable than a “purely human” but poorly thought-out ramble.
The obsession with detecting AI usage by style rather than evaluating creativity by substance feels like a distraction from what readers actually want: informative, engaging, authentic human perspectives on topics that matter to them. It was not until a few months ago did I opt to monetize my articles.
Not all of my writing is AI assisted, I never wanted to be a Henry Wadsworth Longfellow even though he had a significant role to play in my history as one of the supporters of my great great grandmother, Mary Walker. I just want to tell my stories, and ChatGCP specifically Claud from Anthropic helps me do just that.
The Training Data Panic Needs Context
The hand-wringing about neural network training on existing content strikes me as fundamentally misunderstanding how both AI and human learning work. AI learns from patterns across millions of texts, much like humans absorb language, style, and knowledge through a lifetime of reading. We don’t accuse writers of plagiarism for having read books and been influenced by other authors.
This “witch hunt” mentality around AI training data feels over the top when we should be focusing on the actual value being created for readers.
What Medium Should Actually Focus On
Instead of trying to police AI usage, Medium should double down on what makes the platform valuable:
Substance Over Style Detection: Judge pieces by whether they offer genuine insights, useful information, or authentic personal experiences — regardless of the tools used in their creation.
Transparency Without Shame: Encourage writers to be open about their process, including AI collaboration, without treating it as a confession of creative failure.
Accessibility as a Feature: Recognize that AI tools can democratize writing by helping people with various challenges participate more fully in written discourse.
Reader Value Over Purity Tests: Focus on whether content serves readers rather than whether it passes some arbitrary “human-only” standard.
A Practical Path Forward
Medium three principles are solid, but I’d suggest reframing them:
- Human stories and experiences first — but don’t gatekeep the tools people use to share them
- Protect writer incentives — including the incentive to use whatever tools help them communicate effectively
- Reader agency — let readers judge value for themselves rather than pre-filtering based on creation method
The collaborative approach I’ve documented in my writing with AI shows there’s a rich middle ground between “pure human” creation and “AI-generated slop.” This is where some of the most interesting and valuable work is happening.
Let’s put aside the pearl-clutching and focus on what truly matters: helping people share their ideas, experiences, and knowledge in ways that serve their readers. The tools are just tools — the human behind them is what makes the difference.
Brian Connelly writes about technology, cryptocurrency, and the intersection of human creativity and AI tools. His work has been never been featured in any publications exploring the collaborative potential of human AI partnerships. Howeve is is still some cool sh$&.
