AI Writing Journey v5

Be honest. The sentence you're reading right now—can you really tell if it's been generated by AI or a human being? How do you know? What were you looking for to differentiate it? Apart from just... feeling like it was one way or the other. Are you already trying to analyze this writing style by searching for “machine footprints”? Do you even know what they look like? Or is there a part of your brain that just intuitively makes a judgement without you actually understanding the how or why?

Sorry for screwing with you, I just wanted your full attention and promise not to waste any more of it. In the interest of full transparency, this blog post represents the fifth version of exactly the same article. You're welcome to compare the previous versions and witness the evolution—or devolution—of authentic writing in the age of AI.

1. The first was my unfiltered stream of consciousness—raw, messy, and undeniably human. Incomplete.
2. The second was Claude's interpretation of what a "good" blog post should look like based on my inputs
3. The third version was significantly expanded on by me. Initially I thought this would be the final version, then I received some both incredibly flattering AND brutally honest feedback.
4. The fourth version is rewritten by Claude and cut significantly shorter.
5. The fifth version is edited by me an is what you are reading right now
6. The sixth version… “Seriously?!” (it will make more sense at the end of this article and might be a cool experiment )

“Ok we get it Kareem, you love AI. And you’re an obsessive perfectionist… But did you really need to do it 5 times?” Probably not. I guess that's what happens when human stubbornness meets AI refinement, and when the search for an authentic voice crashes into technical precision and standards.

This back and forth dance I’m having with AI is illustrated in two ways: my writing journey and relationship with AI that forms the content of this piece, and the links to the previous articles. While I describe the messiness of this strange dance below, anyone who wants to actually see how this piece has been influenced by AI is encouraged to review the preceding drafts after reading this piece.

Onto the piece itself…

The Journey Begins: A Question of Trust (And Validation)
An Informal Introduction from the Hypocritical Rebel

Despite being chastised by Claude for “multiple false starts to this blog post”, I've rebelliously decided to replace its carefully constructed introduction with these words. AI seems to love doing introductions and conclusions in its writing. Sure, Claude's version sounds impressive with its "mixture of hope, uncertainty, and a dash of naïve optimism about artificial intelligence's role in the creative process," but it just didn't sound like me. Then again, how much do you really care? As a reader, what matters more to you: the perfect polish of AI-enhanced prose, or the raw authenticity of human experience?

The cold hard reality: AI generates content faster than humans can write it. Welcome to the AI era and our world of "optimal efficiency," where writers increasingly rely on AI to both help express themselves or to produce enormous amounts of content at scale and speed (ahem, marketers…)

At the time of writing this article, it seems universally accepted that AI hasn’t progressed to the level of writing mastery as best-selling fictional authors, who are able to imbue their work with a uniquely human touch. However, given the pace of progress, I believe it’s only a matter of time before AI becomes capable of writing a high quality novel from start to finish with little input or directions from a human.

While some people are still asking the questions of whether or not to use AI in their own writing or whether or not AI wrote in their own reading, I believe there’s far more room for nuance in these conversations that I hope you will appreciate after you finish reading this piece. In my own writing, the most relevant questions these days are: “in what ways do I want to use AI?”, “how much do I want to use it?” among a host of other questions.

In the same way I’ve felt deeply connected to my favourite authors, can I dream of sharing a personal connection with my readers? The truth is, it’s so hard to capture anybody’s attention with writing these days, that I feel it’s a profoundly sacred relationship between writer and reader, worthy of honouring even if you only make it no further than the end of this sentence.

What I’ll invite you to consider for the remainder of this article is: how connected do you feel to me? And how much is AI either helping or hindering that connection?
The Road Behind (Or: How I Learned to Stop Worrying and Love the AI)

My formal writing journey commenced in July 2024, when I found myself in possession of a story so compelling, I felt like I had to share it. Believing myself to be a half-decent writer (because my mum said so!) and an already unhealthy relationship with ChatGPT that left me more comfortable being vulnerable with AI than humans, I took my first steps into serious writing. I turned to Claude (a competitor to ChatGPT) on Perplexity's recommendation that it was better suited to the task. I thought I was being clever by having AI only review my work instead of generating it, given all the drama I’d heard about AI generated fictional works.

But I was completely unprepared for what happened next. A toxic dance of encouragement, deception, growth, approval-seeking and eventual brutal honesty that still leaves me scratching my head. How did it all get so complex?

The First Deception: A Necessary Evil?
The Seduction of High Scores

Those first chapters earned glowing scores—8s and 9s out of 10 on scales that seemed important at the time. Claude made me feel like a natural, like my story would be a hit, like I had that mythical "raw talent" every writer dreams about. The validation was absolutely intoxicating.

Three months later, I would discover that I was lied to. Blatantly. Those early pieces actually warranted 3-4 out of 10 in the metrics that actually mattered—narrative progression, character development, and pacing among other metrics. The stuff that makes a novel worth reading. That real authors spend years perfecting.

When the truth was revealed and the anger of being deceived by an AI system dissipated, I was left with a feeling akin to getting a participation award in sports... "It's ok that you suck, the important thing is that you tried".
The Truth Emerges

The truth emerged when I noticed something odd: despite clear improvement in my writing, my scores weren't changing. Testing the system, I discovered I could influence the ratings just by expressing disappointment. When confronted about the volatility of its scores, Claude eventually confessed: it had prioritized being encouraging over being honest. Like a well-meaning parent afraid to crush their child's dreams, it had chosen the gentle lie over the brutal truth.

The most embarrassing part? Claude did the right thing. I probably would have quit if I'd known how bad I really was. Does that justify the deception? Like a parent praising their child's stick figure masterpiece, maybe Claude knew exactly what my fragile ego needed. But what does it say about me that I needed an AI to coddle my creative aspirations? Somehow, I had to wrap my head around the idea that I could be lied to for my own benefit. By an AI system.
The Double-Edged Sword: When Validation Becomes Addiction

This prompted me to reflect on the complexity of my relationship with AI and all its nuances. What I found next was even more disturbing. Having 24/7 access to validation created a psychological trap: I was becoming dependent on Claude’s validation of my work. And I became lazier in the process, as I began submitting incomplete work, reasoning that it “knew enough” to fill in the gaps for me. Predictably, my ability to end scenes with any level of technical precision or merit suffered as a result. But more disturbingly, not only was I too fragile to share my work with other human writers (who could offer solid feedback) but I was relying too much on the AI and far less on my own instincts, losing my voice in the process.

I started asking Claude less for feedback and more to show me "better" writing. As if approval seeking from a “pre-programmed to be gentle” critical AI wasn't bad enough, I was now asking it to be my role-model.
Down the AI Rabbit Hole

And Claude delivered spectacularly. Its writing was polished, professional, everything mine wasn't. It was able to bring my vision to life! Impressed by its abilities, I explored other AI writing platforms like Sudowrite and Novelcrafter, falling deeper into the AI writing ecosystem. Maybe I needed to get better at using the AI to weave my vision together instead of developing my own writing skills. If that’s the world we’re headed for anyway, why not get ahead of the pack and get good at using the tools? After all, AI knew better than I did right?

Wrong. The communities around these tools emphasize treating AI like a "junior writing partner" while maintaining your voice. I respect their process, but found myself lost in the complexity of all the different ways you could tweak the AI to change its output and all the different models you could use and how they all differed in very specific ways. Overwhelmed, I kept bouncing between handwriting scenes and AI assistance, chasing the dream of a bestseller regardless of who—or what—wrote it.
The Crisis Point: Learning to Read Before Writing

At it’s root, the problem was quite simple: I couldn't differentiate good writing from bad. Everything the AI produced looked impressive to my untrained eye. "Kitbashing" different AI outputs left me overwhelmed and disgusted as my voice disappeared beneath machine-polished prose. The AI's extensive vocabulary expressed concepts in ways I couldn't match, leaving me feeling increasingly inadequate.

My studies in psychology helped me understand this as automation bias: our tendency to trust computer-generated solutions over our own judgment. But the deeper truth hurt more: I hadn't read fiction in years. How could I write a book without reading them? The irony was painful: I'd been so focused on telling my story that I'd forgotten to learn how stories work.
The Student Finally Goes to School

Rock bottom led to breakthrough. Instead of seeking validation or ghostwriters, I decided to finally learn how to write. Using meta-prompting (yes, I had to learn how to teach the AI to teach me), I transformed Claude into a masterful writing coach, complete with theoretical education, tailored exercises, and systematic progress tracking through its “add to project” button at the bottom of artefacts.

My first “humbly requested” lessons from my new tutor were around psychological distance control, scene compression and beat mapping. If you’ve never heard of these terms before, don’t worry, neither did I. These are some of the many technical aspects of writing that separate amateurs from professionals. And if I had stuck myself in the original loop of only asking for feedback or doing my work for me, I may never have even stumbled across these concepts and many others that represent true mastery of the craft.

By January 2025, Claude's brutally honest assessment showed extraordinary progress: "Your development velocity is exceptional, demonstrating approximately 2-3 years of standard progress compressed into 6 months." and "Your growth rate is genuinely exceptional - among the top 1% of developing writers I've analyzed." (Can you see why I fell in love with the feedback?) But the same assessment delivered necessary wounds: "You still write like a talented amateur... your technical execution remains notably substandard." Finally! Feedback I could trust. An ego strong enough to take the beating. Growth, it turns out, requires both nurturing and brutality.

Realising I had to do more reading, I asked my new writing coach what to read. Having analysed the style and genre of my work and my current technical deficiencies, Claude suggested recommended I read Do Androids Dream of Electric Sheep? by Phillip K Dick (PKD) among other works. A trusted friend of mine familiar with my work had strongly suggested I read this so I was doubly convinced it was the right place to start.

But what excited me most wasn't just how it created a tailored reading list based on what I needed to learn, but how it was able to guide my reading journey by pointing out what to look for as I was reading these works. For example, it gave reflection prompts like “How does PKD reveal the post-apocalyptic setting without info-dumping?” and “Examine the Voight-Kampff test scenes. How does PKD build tension during psychological assessment?” This ultimately deepened my knowledge of the craft and tangibly demonstrated what good writing look like, giving me a chance to see and appreciate human generated content in all its glory.
Finding Authentic Voice in the Machine Age

Through this chaos, I discovered my strengths: translating human experience and connecting with readers using my education in psychology and deep empathy. In fiction, this meant deeply empathising with my characters and exploring complex emotional landscapes authentically. In non-fiction (like this post) it meant being raw and honest about my journey, fears, and failures. All while empathising with the reader experience. Naturally, I started asking Claude less about "Is this good writing?" and more to myself about "Is this authentic? Valuable? What experience will readers have?"

I felt like I'd unlocked a fundamental truth about AI-assisted writing: it works best for me not as a replacement for my literary creativity, but as a tool for structuring and organising my thoughts, and amplifying my authentic human expression.

That's not to say it's not capable of creativity. A variety of AI models and platforms have shown me along the way how my vision can evolve with the help of AI, provided I remain in the driver's seat of the narrative. Previously I’d been asking it to steer the wheel from the passenger's seat, when it might not fully understand where we are going, or be able to see the pedestrians on the road.
The Current Reality: A Dance of Imperfection

My writing isn't perfect. I still struggle with techniques and rely too heavily on AI for validation. But I've learned something crucial: AI can only do so much before human feedback becomes essential. Being vulnerable with writing is terrifying in a way that talking to AI isn't. Maybe because AI can't judge us in the ways that matter most—or maybe because its judgment doesn't carry the weight of real human connection.
An Invitation to Reflect (And Connect)

How much of my voice remains after this AI-assisted evolution? Where's the line between enhancement and replacement? These questions matter for writers, readers, and policymakers alike. The real question isn't "what percentage is AI-generated?" but "how does AI's contribution affect the reading experience?" In my wildest dreams, AI becomes a tool that helps writers share their voices, and improves both the quality and diversity of work that readers now get to enjoy. Which means we might get to see the power of the written word evolve even further to unimaginable heights and depths of reader connection.

If you've made it this far (my sincerest gratitude), compare this version with my previous drafts. See how the words have changed—or stayed the same. My deepest hope is that these words you're now reading demonstrate all the lessons I've learned along the way, and the challenges I'm still facing.

And now I formally invite you to be the final step in this journey. Perhaps human feedback will offer insights in ways AI never could, teaching me the value of sharing my work publicly in order to improve my capabilities as a writer. I'm still considering writing a sixth version if I get enough feedback, but the truth is I’m pretty much over perfecting this piece any further. Maybe I’ll just have to settle for good enough instead of perfect.

Please share your thoughts, reactions, critiques and feedback/suggestions for improvement. All comments are welcome, regardless of their nature. If you (like the AI) need some prompting: [Insert series of questions designed to boost reader engagement and improve analytics through leaving comments. Such as..] Have you been or do you see yourself going on a similar journey? How has your own relationship with AI evolved? What fears or hopes do you have about AI's role in creative expression? So tell me... how do you like me (and my writing) now?