About Fiction DNA
Hi! I’m Kyle. I don’t have an MFA but I like all types of Art and I like learning about them on my own. Often when I read a Science Fiction story, like Story of Your Life by Ted Chiang, I’ll wonder - did he invent the idea of an incomprehensible alien form of language, or is this story the evolution of an established concept? China Miéville is doing something like it in Embassytown, where the aliens speak with two mouths at once and can’t say anything that isn’t true. I found one of those books because of the other, and that’s exactly the kind of thread I want to be able to pull on. I ask myself this question about 50 times any time I read a Neal Stephenson story, and I spend tons of time on Wikipedia trying to figure it out.
To help in my own writing I have a stack of books that discuss plot structures, narrative techniques (the Elements of Fiction Writing series) or the practice of writing itself (On Writing, Stephen King) … in addition to the mechanical things like The Elements of Style and The Craft of Research. I like to dissect things and then try to build something from the pieces that make up good storytelling. It’s just how I think about things.
I’m a software engineer who wants to be an artist (looking at you Paul Graham, Hackers & Painters), which led me to creating Fiction DNA so I could more easily pick apart all of the things that my brain wants to when I’m thinking about books I’ve enjoyed or my own writing. I hope it’s useful to people who care about similar things.
But what about the use of AI in creating all of this? Yes dude, most of the analysis-type content is AI generated. I am using Google’s Gemini models with some finely tuned prompt-engineering to help with that. To me this seems OK, to you it might not and that’s why I want to mention it. In my day job I am focused on how to implement “Ethical AI” systems, and this app is an example of that, based on my definition - so here it is:
- Humans make decisions: The LLM can never act autonomously (modify data), it can only suggest things and a human needs to review it.
- Transparency: Any content generated by an LLM must be labeled as such.
As you can imagine, lots of details flow from those two elements but this is an About page not a tech spec so I’ll leave them out.
A couple more things I’ll mention about AI to round out my thoughts and, maybe, give you a lens to consider your own.
- I think LLMs are fundamentally bad for society. The business model of the big companies making LLMs is, fundamentally, replacing human jobs. If they say otherwise they are lying.
- LLMs are disastrously large energy consumers. This is bad for the environment, for communities that suffer the scourge of data centers, and feels like it will lead to a Matrix-like world where humans are all batteries to power the AI grid… honestly, it feels that way sometimes.
- LLMs have incredible promise in research and science and there is potential to use them as tools in ways that create good for humans. This is a belief and hope - I hope to see more evidence of this.
So, why do I use LLMs? … I don’t know, man. Why do I eat meat when I know it causes suffering to animals and huge ecological damage? Why do I drive a car when I know it’s contributing to world-altering climate catastrophe? I guess I just pick my battles. I like people - what they think and what they create. I think art does change the world, if indirectly. But ultimately, I don’t know how to function as a human under a cloud of dread and Art is the sunshine my soul needs.
Best case scenario with Fiction DNA is that some people will enjoy exploring ideas and discovering some stories to read. Worst case scenario is that only I will. That’s not so bad, so that’s why I built it. I hope you enjoy.