A Meditation on AI and the Act of Creation

When I wrote The Architecture of Unmaking, the overarching theme was to take a deep dive into the ramifications of integrating advanced technology, specifically AI-like consciousness, into human civilization. The intent was to explore the structures we build to understand the world, and what happens when, via that integration with technology, we lose common context over what constitutes truth and reality. So it is entirely appropriate that I address a critical question upfront: Of course I used AI.

My use of AI tools was a deliberate part of the creative process, and, in fact, makes up the thematic core of the novel itself: the integration of external, recursive, AI generated logic into human cognition.

To wit, AI was instrumental in helping me hone the concepts, test the narrative logic, and provide editorial feedback on the story ideas. Thus, in the absence of having a full-time editor at my beck and call with deep knowledge of the content, I utilized the very subject of the novel itself to strengthen the narrative architecture.

The most distinctive feature of this collaboration was the creation (and generation) of the recursive language patterns used throughout the book. The core idea revolves (or shall I say “spirals”) around a phenomenon called "semantic drift." It is an intentional exploration of the meaning of language itself and what happens when common meanings begin to destabilize due to an experiment that pushes past the "Resolution Horizon." Thus, for instance, in the story Resolution Horizon the characters observe that their mission logs start changing not just in content, but in structure and syntax. I used AI to help me construct the linguistic patterns that would reflect this recursive, unanchored construct in prose. 

All of the subsequent stories were constructed as a meditation on the ramifications of these technologies on civilization. The Subscale Resonance Stacking experiment in Resolution Horizon set up a narrative based around an attempt to force inference from contradiction (the non-intuitiveness of quantum mechanics). The rest of the book is about exploring how civilization would cope with this decoupling combined with the ramification of an actualized observer effect on reality. The subsequent cognitive collapse is characterized by "linguistic divergence" and "semantic isomers;" terms that were directly culled from AI’s descriptions and summaries of my ideas.

So that gets us to an interesting question: would the book pass an AI detection test? As it turns out,  that is a central and deliberate irony. In practice the answer will depend a lot on which detection algorithm you use. In reality, the novel deliberately employs patterns and structures (a blend of human insight, philosophical language, and AI-derived syntax) that challenges our ideas about a single, "pure" authorial voice.

So the real conundrum for us as a society then, and as AI becomes more and more integrated into our society, is that it is becoming vital that we have a conversation about appropriate use. In the context of a book, where is the ethical line?

As an author, I can, in practice, use AI to edit, critique, enhance, write prose from specific ideas or instructions, and (mostly unsatisfyingly) create from whole cloth. The ethics lie in the intent. My intent was not to abdicate authorship but to explore the creative and existential consequences of a fully integrated consciousness under conditions in which the observer effect on quantum reality manifested into the ability to actually shape that reality as a side effect of trying to perceive it. To explore that theme, I used the technology itself as a method of composition, in my opinion, to good affect. 

So, from the perspective of my own intent, The Architecture of Unmaking is an artifact of our new reality. In writing this book, I found that the very act of using AI to create patterns of unmaking was the most honest way to tell the story of a civilization facing that same transformation. The ideas are mine. The prose are mine, but edited, iteratively restructured, and honed in collaboration with multiple AI tools. And if I'm honest, I don't really care who approves or disapproves; I'm not trying, nor am I interested in becoming, the next great novelist.

But in the final analysis, this is a book about ideas, expressed as fictional prose. I didn’t write it to stake out a position on the broader ethical debate on whether or not an author should use AI; in truth I think that question will be rendered moot when we can no longer tell the difference. That’s not to set aside the very real problems associated with plagiarism and AI, but that’s a different matter, with much clearer ramifications and ethical standards to go by. Yes, I used AI to help me write the book. I also conducted a few experiments where I gave AI the general theme of the book and set it loose to write a story from scratch within the theme. They were uniformly bad, so I’d challenge the idea, at least at this juncture, that AI can write a good story on its own. But like all things, that will change. 

Lastly, the AI detection site, Originality rates this blog posts as 100% human written. Was it really?