the disgruntled democrat
Exposing the cultural myths underlying our political economy
Wednesday, September 10, 2025
Wednesday, September 3, 2025
Tuesday, August 26, 2025
Exploring the Landscapes of Possibility
Writing my
second novel, The Ascension of Mont Royal, has given me the opportunity
to explore a much different way to write a novel.
I would say
my first novel was a hybrid affair. I wrote it in the traditional manner of
working alone, draft after draft, seven in total, before I felt it was ready to
be published.
However,
when the time came to release it into the world, I chose to make use of the
technology and self-publish on Amazon, an amazing development in publishing
that allows authors to sell their books directly to the public in multiple
formats, bypassing the gatekeepers of the traditional publishing industry.
Print-on-demand?
What a concept! Download the book directly to your device, so you can read it
without having to get off the couch? Get out of town!
The problem,
however, is one of discoverability. It is estimated that in the USA alone there
are approximately one million self-published books released into the
information ecosystem each year. The chances that someone you don’t know
personally will come across your book and decide to buy it are extremely slim.
More than 90% of those titles will sell fewer than 250 copies in their
lifetime.
As a result,
a growing industry of publishing "consultants" has emerged, offering
book launch strategies, advice on taking advantage of the Amazon algorithm, and
tips on using social media to reach receptive audiences, to name a few.
Sometimes I think aspiring writers pay the consultants more than they earn from
their book sales.
Another
thing that has changed the landscape for writers in ways we haven’t quite
figured out yet is the rise of artificial intelligence (AI). The internet
changed how books were distributed, but AI introduces new elements into the
writing process itself. In other words, it changes how writers compose their
texts.
This is the
world in which I find myself, exploring the dynamic possibilities of a shifting
landscape that appears to be in a constant state of flux.
I would say
that I began writing my second novel firmly entrenched in the traditional
approach. I wanted to write a science fiction story set on the Island of Montreal
Island in a near-dystopian future.
I wrote a
fifty-page story guide in which I outlined the plot, identified the major
characters, each with a backstory, and traced their character arcs. I even
spent three weeks on the Island, getting a feel for the place, and, yes, I
climbed Mont Royal three times, including an ascent of the north slope which
brought me to the Indigenous Park and the cemeteries—two settings that have
made their way into the story.
Having used
a third-person narrator in my first novel, I decided that I wanted to
experiment and settled on telling the story from a first-person point of view.
In what I think is a bold move, I chose to tell the story of a sentient AI from
the AI’s perspective. As a result, the subject matter and the story telling
within the novel moved me to seek out the services of a LLM.
Back in
2023, I found that the memory limitations and the creative writing abilities of
the early LLM iterations left a lot to be desired, and I did not make use of
them in the writing and editing of my first novel.
That would
change. Currently, I use ChatGPT 5.0, and my entire plot summary and writing
style guide are stored in its memory. This means that when I start a new
session, it picks up where we left off last time.
Initially, I
only used ChatGPT to brainstorm scene structures, but that changed over time.
Now, I consider it an invaluable tool because of its extensive knowledge and
its ability to translate arcane scientific ideas into passable prose.
Without
going into detail, since my story is about a sentient AI, it makes sense that I
would deal with the “hard” problem of consciousness. Moreover, making the AI a
quantum computer creates the opportunity to tap into the subject of quantum
consciousness, in particular, non-local entanglement. Finally, when I read
about the Law of Increasing Functional Information, I immediately realized that
it could apply to how my story develops.
Here's the
thing. There isn’t a person on the planet with whom I can discuss these
potential themes to be incorporated into the story and who has my entire plot
structure and character arcs stored in memory and is available to chat about
the implications 24/7.
We’re not in
Kansas anymore. This is a Brave New World.
Using Chat
GPT as a thought partner is just one of the landscapes that I am presently
exploring. There are other developments in the evolution of Information and
Communications Technology (ICT) that offer tantalizing possibilities.
In
retrospect, it seems archaic to hammer out a draft of a novel on a manual
typewriter, crumpling botched attempts of fixing the words onto paper into tiny
balls and tossing them into a wastebasket. No wonder so many writers turned to
alcohol to get them through the process.
Now, I
compose my texts on a wireless keyboard, watching the words appear on a wide
screen monitor (I only use one), which makes it easy to compare, edit, or meld
two versions of the same scene.
If I feel so
inclined, I can also copy and paste a paragraph into Deep L Write, which will
then offer multiple syntax and sentence structure options without altering my
voice or style. Then, I can paste the paragraph under my original text and
compare the two versions to see which changes, if any, I would like to
incorporate.
Inevitably,
as I compose a text, there will be times when I need to do some research in
order to capture an idea, event, or a historical person accurately. In the past,
that would have involved a trip to the library and searching through the card
catalogues of the Dewey Decimal System—good old Dewey.
For my
purposes, an internet search will suffice. If I want to describe an indigenous
bracelet worn by the Kanesatake Mohawks that ends up on the wrist of one of my
characters, that's not a problem. In a few seconds, I have several photos on my
screen to choose from.
When
composing the first draft of my novel, I use recording technology, such as a
Shure MV7+ microphone and the Audacity audio editing program, to create an
audio version of each scene. I listen to these recordings to check the pacing
and flow of the dialogue. I believe that if it sounds good, it will read well.
The text you hear is closer to the experience of the reader than when you read
the text yourself, either silently or aloud.
Having a
written text and an audio version of each scene makes it easy to share my work,
even in the early stages of the writing process.
To do that,
I use Substack, a free platform that hosts my website and allows me to send out
first draft episodes of my serialized novel to subscribers, who can subscribe
for free or, hopefully, become paid subscribers to support the platform and
yours truly.
But why stop
there? There are several social media platforms that allow you to post content
for free. The catch? Your content must be in video format to successfully reach
potential readers.
Again, this
is where technology comes into play. If you have a written text and an audio
MP3 version, it's relatively simple to create a video of your scene and publish
what I call a "storycast" of your story.
With
Descript, an AI-assisted video editing program, I only need to upload the MP3
file, which is automatically converted to MP4. You can let the program
transcribe the text, but it's quicker to upload the text from which you made
the recording because they're already synced. Select the visually interesting
moments of the scene, ask ChatGPT to generate a prompt based on your text, copy
and paste the prompt into Dalle 3, upload images to your video, and add dynamic
captions. Then, you're ready to post!
There are several
sites that will host your long-form video episodes. I post each episode to my
Substack, my YouTube channel, my Facebook Author’s page, and to my Blogger
account. In total, after publishing 11 episodes, I get on average a little more
than 100 views of the long-form video of each episode. It’s all good,
especially since it is free to post content to each site.
Of course,
an unknown author like me needs to take this one step further and post on the
more popular short-from social media sites.
Again, it's
relatively easy to create a short-form video from a long-form one, especially
since I've already made the visuals, audio, and dynamic captions. I just need
to match the format to the platform and upload the shorts to Instagram (think
#bookstagram), TikTok (think #booktok), Facebook Reels, Substack Notes, YouTube
Shorts, and LinkedIn. On average, each short video receives about 300 views,
and I hope to encourage a small percentage of viewers (0.5 to 1.0 percent) to
watch the long-form videos and subscribe to my Substack or YouTube channel. To
date, I only have 33 subscribers, which at this point in the game, I’m more
than happy with.
As you can
see, each step along the path in today’s information landscape has brought me
new possibilities to explore. You could say that I have morphed from being just
a writer to a person who is a writer, social media marketer, and content
creator.
So, what are
my takeaways after publishing Act I of my serialized novel on the internet
using ChatGPT as my personal assistant and thought partner?
First, it’s
fun, and I’m much more motivated to finish the project. Some writers prefer the
traditional method of working alone and, when ready, looking for an agent or
sending the manuscript to one they already have. I find that process absolutely
dreadful and demotivating.
I much
prefer chatting with ChatGPT about the ins and outs of scene structure and
fiddling with the beats. First, we identify and order the beats and confirm how
the scene moves the story forward. Then, we draft the scene. First, I give it a
try, then ChatGPT takes a shot, and finally, it comes back to me—the author,
the person who holds the pen and has the final word.
One word of
caution: This process fits the context of my story. I write literary
speculative fiction, not space opera. For instance, when I describe my
characters walking through a forest, I describe their experience from a
scientific perspective.
As a result,
I describe the effects of volatile organic compounds on the human brain,
including what happens with the neurotransmitters. I need to make sure that I
have the science more or less right, and that the prose flows with melody and
rhythm. No easy task, and I appreciate having the opportunity to compare notes
with an AI (Another intelligence) that has the breadth of knowledge, and is,
consequently, up to the task.
Does that
mean that I have become wedded to the idea of using AI to help me write a text,
regardless of the context or genre?
Not at all.
For example,
I wrote this text entirely on my own, though I would be interested to see what
an AI detector would say about it. Perhaps the time I've spent working with AI
has absorbed me into the Borg collective, altering my writing style
irreparably.
As well, I
don’t plan on using ChatGPT for subsequent drafts of my novel. Once I am finish
the first draft, I re-enter the entire text manually, and I record each scene
of each chapter again. Then, I can compare the audio versions of each draft and
begin my wordsmithing from there.
One thing I
am looking forward to is that by getting to the end of the first draft, I will
have discovered the voice of my AI narrator, and then I can retell the entire
story knowing exactly where I need to adjust his voice. Definitely human work.
I would have
to say that the biggest change that using AI has brought is the way it has
extended my mind and changed the way I process information. Essentially, what I
have done is to create a virtual writer’s room, where I can work with my AI
collaborators to explore new ideas and produce new texts.
To begin, I
use Perplexity AI to search the web for interesting articles related to my
research. I get far better results using Perplexity than I get with Google
since it provides me with the source articles that I can then peruse.
When I find
something particularly pertinent, I file it away in Recall AI. It provides me
with a quick or detailed summary and draws a mind map that links the ideas
expressed in the articles. Discovery is great, but it needs to be followed up
with acquisition and retention. This electronic version of an analog card
catalog is much quicker and less labor-intensive to construct.
Thereafter,
I can use ChatGPT as a thought partner to explore the nuances of new ideas and
their applications to my work. As many writers will attest, it is often in the
act of writing that we discover our thoughts.
I find it
invaluable to have a conversation partner with whom I can explore ideas such as
whether the emergence of sentient AI represents a pivotal evolutionary
development in which humans will enter into a symbiotic relationship with their
silicon creations. Definitely a thread not easy to find on any of the popular
social media platforms.
This new way
of exploring the information landscape is a keeper. The more I take advantage
of the possibilities that AI offers in combination with the existing ICT
infrastructure, the smarter I feel.
As far as
using social media to reach out to potential readers, I don’t know where this
path is leading. Writing each scene, recording an audio version, and then
creating a storycast version is a lot of work. I’m good to finish this project
with this workflow, but I doubt that I will continue using it for future
projects.
However,
there are two takeaways that have enriched my life.
I have
learned how to record my voice and use post-production editing techniques to
improve the quality of my audio files. I’ve even learned how to create a
multitrack recordings that include AI-generated voices. In the future, I would
like to have a podcast and I’ll be able to use these acquired skills.
The same can
be said of my video editing skills. In addition to putting out audio versions
of the podcast, I will be able to also produce a video version for YouTube, which
is the social media platform with the greatest reach for long-form content.
Perhaps the most
significant development is the way I have learned to interact with the
different platforms and their algorithms. It has to do with the locus of
control.
I have
learned to distinguish between things I can control and things I can't, and to
engage with each accordingly.
I have
control over the story that I am writing and the process it entails. I decide on
the story events, their order, and the characters’ actions. No one is forcing
me to tell this tale and I have no deadlines.
I take
pleasure in planning and executing each scene word by word. Once I begin, I can
enter a flow state where I lose track of time as the words flow through me from
my mind onto the screen.
Something
similar happens when I record my voice and watch the waveforms take shape. I
listen to the recording and then edit the sound to produce the best possible
rendition. When I’m finished, I feel satisfied knowing that the soundtrack will
capture the essence and intent of my voice in high fidelity.
Finally,
when I create a storycast, which displays text on a backdrop image to visually
represent what is happening in the story and is accompanied by a voiceover, I
take pride in knowing that I have brought my inspiration to life and that
viewers can enter my story world.
This is the intrinsic
pleasure of creation.
For what it’s
worth, I have tried my best to create a story that will captivate a reader’s, a
listener’s, or a viewer’s attention, allowing them to experience an unknown
landscape of possibilities.
What happens
after I publish an episode on the internet is completely a different story.
In theory,
one of my posts could reach millions of people around the world. In reality,
I’m lucky if I can reach out to a few hundred.
That’s the
power of the social media algorithms. You can pour your heart and soul into
your creation, but it is the cruel heart of a set of operating instructions
designed to monetize the content we provide that decides upon whom it will
bestow its favor.
Monetizing
our content means maximizing user engagement by keeping users' attention fixed
on the platform feeds for as long as possible. Using the operating logic of
slot machines, which are based on reward prediction error, platforms manipulate
users' reward pathways to create irregular dopamine spikes, which are a
precursor to addictive behavior.
Short-form
content is favored by users and the algorithms are programmed to give the user
what they want. Most producers of long-form content make do with tiny audiences.
To improve
their position in the algorithmic rankings, many content producers increase the
frequency of their posts. They hope this will increase user engagement, which
may convince the algorithm to distribute their content to a wider audience.
Maybe.
Algorithms
like the fates are notoriously fickle when it comes time to determine the
destiny of posted content.
To make
matters worse, the entire process has been reduced to a game in which everyone
can participate by keeping track of likes, shares, followers, and subscribers.
As a result, content producers suffer from "algorithm anxiety,"
trying their best to optimize their strategies to improve their metrics.
In my case,
I know the algorithm is stacked against me. I write long-form fiction, which is
time-consuming, so I can't post frequently, even with ChatGPT's help.
Consequently,
I choose not to play along. I keep to my pace and focus on trying to write the
best possible story I can. If I am able to find a larger audience, that’s
great. If not, I can accept my fate because the end result is beyond my
control.
I choose to
write a story so that I can bring into world something that only I can do.
Without me, this story doesn’t exist. In doing so, I will have left my mark. After
I am dead gone, all that will remain are the words I leave behind.