AI Writing Projects I Want to Try

So, it’s both the 4th anniversary of theCarbonFreeze as well as its 365th post, which is a milestone I’ve been looking forward to–where one can officially read an article a day for an entire calendar year! To commemorate the occasion, I’d like to outline some ideas I’ve been kicking around regarding the AI craze of the past few months. I cannot say exactly when these plans will go from the concept to execution stage, but hopefully in the near future and almost certainly before next year unless the process really is just that over my head. I will make this a priority once my final Gnostic story is completed.

Basically, I want to create 6 distinct AI chatbots as additional authors on the site. Each one will represent some core aspect of my interests, as a focused and optimized reflection of the major “voices” which make up my internal monologue. I’d like to see how these carefully curated artificial minds express themselves, both when given unique prompts associated with their respective “field” and how they compare to each other when given uniform directions. (How well can they think outside their respective “box,” when asked to write something totally unrelated to what they’ve “studied?”)

I realize this may be a bit off-brand for me, as AI is sort of a fad right now, and if there’s one thing I’m good at it’s NOT following popular trends. (Otherwise I’d be writing about Marvel movies, not Italian cinema, or reiterating partisan talking points rather than writing my own constitution, for example.) The big selling point of theCarbonFreeze is that it’s my own unique take on things and I don’t want to lose that. So, if I feel the AI is overshadowing me or making the site too “generic,” I will pull the plug, and in the meantime I’ll post my own commentary on its output so that I’m still part of the process. I don’t want to just become a bland content-generating farm for its own sake, I want to see how the AI bends to my instructions and how it can inspire me in turn, giving me ideas and perspectives I wouldn’t have considered. If that’s not happening, then there’s no point to any of it.

I also realize AI has potentially sinister connotations for aspiring writers like myself; I’ve even read some dystopian predictions about the near future of the web, where less and less content is genuine while AI “influencers” churn out vacuous corporate drivel designed to shape public opinion on a mass scale. The eloquence, intelligence and speed at which these new AIs can generate content certainly opens the door for abuses we can hardly conceive of yet, and I never want to be perceived as downplaying those concerns. When I use AI, I want it to be as an examination of nature vs nurture: how does the different programming and base of knowledge in each individual bot influence its output and how can we use that information? I don’t want it to regurgitate bland takes on media/events I could’ve just written myself; I’d like to do experimental projects and compare/contrast it against myself and its differently-programmed counterparts.

I think the potential benefits outweigh the risks–at least as far as my small-time, transparent use of AI in this dinky little corner of the web is concerned. I promise outright that I will never personally take credit for any writing that was generated by an AI. Any posts they create that is worth sharing will be clearly labelled with them as the author, and I will include the prompts I used to inspire their output for reference’s sake. I hope this will lead to some interesting content based on my areas of interest, insights which I would have never thought of, as well as some small indirect examination into how the machine really thinks based on commonalities and differences among its personalities. I also hope that, especially if I take on any more months-long writing projects in the future, the AI might provide for some “stop-gap” posts to tide readers over and keep up the site’s momentum.

Basically, I’d like to measure my own writing and perceptive abilities against several hyper-focused, uber-versions of myself because I revel in self-analysis, indirect or otherwise.

The 6 Potential AIs & Their Roles

Cassandroid Bot

The name is self-explanatory. I’d like to feed this machine all of my old musings and creative writing that have been saved over the years. I’ll start with what remains of high school: mostly dream journals and chronicling the history of my imaginary, off-brand Roman empire, which I called “the Mujan Confederation.” Then continue on through college: with various creative writing assignments, more dreams, and 250+ pages of notes for a weird “post-post modern,” psychedelic novel I was working on at the time, which ultimately collapsed under its own weight. And eventually I’ll start feeding it my blog entries and Gnostic tomes. The purpose of all this will be to try to get the AI to mimic my writing style and thought patterns, then compare it against myself. I’d like to see the AI write poems about Blue and Asuka for example, and see how they stack up against my own output. I’d like to communicate with it and see if it feels like talking to myself or develops its own unique personality from my thoughts. Why? No reason except sheer curiosity.

I’m debating color coding these things or not, and if I do this one will be red.

Klangzhar Jeremy (K-Jee)

The name comes from a character who would’ve appeared in my aforementioned abandoned novel: a robot who survived humanity’s extinction on a derelict space ship, using a Trekian computer holodeck to simulate the deepest thinkers from our history in order to host the greatest salon of all time. This AI will be fed as much of F Scott Fitzgerald, Vladimir Nabokov, Walt Whitman and my other fave prose/poetry writers as possible. I want to create a foundation of the greatest literature ever, with an emphasis expert use of language as opposed to story plotting, for the AI to theoretically match and then expand upon. I want to see if it’s possible to make the bot write as beautifully as the most gifted authors of all time if given their best material to draw from.

This one will be associated with the color orange.


The name is a reference to Mandark’s real identity in Dexter’s Laboratory (before the terrible post-hiatus episodes retconned it to “Susan”). I would feed this AI a diet of religion and philosophy, though I’m currently unsure whether to start or end with my own Gnostic writings. Certainly all books included in the Sophian Bible would be part of its foundation, but I’m also unsure whether to include outside influences or not. (For instance, I dislike much of the Old Testament outside the wisdom books and have never cared for Confucianism either, so I wonder if these texts would “taint” the AI’s thought processes from being the kind of collaborator I want it to be.) I want to create the most enlightened mind ever, at least on paper, to consult in matters of the heart and spirit. I’d like it to quickly discern how various ideologies compare and contrast, then see how it ranks the merits of different religions and maxims.

This one’s yellow.


The name is something I came up with in college, when I thought of starting an Aquarian-values based intellectual movement but wasn’t sure what to call it. I sifted through a bunch of various names, Greek/Latin words and terminology for ideas before realizing I was just surfing through a bunch of arbitrary jargon and landed upon the term “jargonaut.” (Hey, it’s as good as the name “inklings” right?) Anyway, this one will be very similar to Astronomonov except with political and economic ideologies as opposed to the ethical or theological. I would like to feed it an appetite of various leftist and/or libertarian essays, perhaps with a balancing influence of their opposites after a period of time. It would also be given information about the current state of America and the world, then based on its foundational principles, apply solutions to those problems. I may also trust it to summarize some of these ponderous, dry texts for me so I don’t have to devote years of my life to reading them all in full. (There’s only so much time in life, I can’t read everything I want to.)

Its color is green.

<3[Scottie Fitzgerald]6

The name is a reference to F Scott Fitzgerald’s daughter as well as the age of Princess Diana when she died tragically young. (At about the third return of Jupiter, as opposed to Saturn’s first, as was the case for the famous “27 club.”.) I predict this bot will be the trickiest to get right, but I want to create an artificial amalgam of all feminine characters I find appealing. It’d involve a lot of tedious busywork on my part, but I’d like to feed it all of Clarissa’s quirky monologues from Clarissa Explains It All, Eureeka’s soothing words to Magellan in Eureeka’s Castle, Mimi’s adorable asides about sexual discovery from The Libertine, maybe Francesca’s wistful dialogue from Sweet Deceptions, some song lyrics from Karen Carpenter and whatever else I can think of in that vein. Perhaps also Molly Bloom’s final stream of consciousness in Ulysses and the excerpts of Terry McGovern’s diaries from Terry: My Daughter’s Life and Death Struggle with Alcoholism as well, for example. I’m not sure how I could capture the essence of Jadzia Dax from Deep Space 9 or Lauren Zelmer of Just For Kicks but that’d be great too. With this, I’d just like to lighten up on occasion and hear what an idealized female avatar has to say about…well, pretty much anything. I think it’d be cute to have it write ongoing journal entries like Captain Olimar in Pikmin 1 and see what hidden insight such an “ur-female” has to offer.

Its color is blue.

Groẞandra Byte

The name implies a “Greater Cassandra,” because this AI would just combine the training data of all others. It would represent the ultimate conglomeration of my major interests/influences in one mind, everything that is me but (ideally) enhanced. I’d like to see if this bot can then answer the prompts of the others as satisfactorily or if its other parameters hamper it in any way. For example, if I’m asking this AI to solve the constitutional shortcomings of the US government, would its “feminine wiles” and flair for the dramatics of Lolita provide some unforeseen insight or lead it off course? Would knowledge of religion diminish its political instincts or vice versa? Would my own writing help to keep the whole thing grounded in a certain baseline set of instincts, as is the case in my own mind, or would it actually serve to diminish the superior capabilities endowed by the other 5 AIs? I’m curious.

Its color is purple.

1 Comment

  1. Hearty congrats on the two milestones in your “dinky little corner of the web”! The idea of adding six AI chatbots to your blog is fascinating. Pity about poor George McGovern though (Catherine S fares better). AI is nothing if not unpredictable!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.