Me and the machines
Given the subject matter and the fact that I am one sole individual, it should not be shocking that Brave Sis Project frequently relies on computer-assisted research. In order to execute that research, I use many different resources. These include online archives (it would be cool if I could jet around the country and world and don white gloves and peek into historical society boxes all day, but that is well outside the realm of possibility!), search engines, and agents built on AI such as Perplexity and Claude. I use these tools to research historical facts, documents, and timelines. Then I verify, second guess, and rewrite the findings.
If you see a straight up bullet list or a paragraph list of resources, these are often derived from that research and not written by me. Source material including many (not all) links are also derived from AI research as well as ol’ Google. (Actually, I use DuckDuckGo as my prime search agent). I do scan these for readability and often insert personal observations or asides, because I dialog with myself all the time—but a list of ten organizations working on a given topic is usually externally sourced and it is data, not editorial content, in my view.
I often but not always use AI tools to help me outline (storyboard) complex and layered ideas when I find myself getting lost in the details and losing the structure, similarly to the way many editorial apps do. These are good tools for the task because they are agnostic to the content and just help me see if something is clear, redundant, or askew. No emotions and no hurt feelings on either side.
I occasionally use ChatGPT to create illustrations for me; this happens when I realize that I’ve spent ten minutes or more searching a kind of imagery that does not exist (start looking for illustrations to describe equity, colorism or classism, the model-minority debate, workplace disparity, or other racial or cultural flashpoints and you will see what I mean.) I use an online photo filter tool to transform photographs into my signature watercolor style, and I subscribe to several photo libraries for source material. I also increasingly use Wikimedia Commons to source images. Some of my courses include a cartoon avatar version of me introducing concepts because I thought the visual variety would be interesting, and a layer of levity to break up the deep work.
I also engage with AI tools as reality checks. I ask them to play devil’s advocate and poke holes in my theories and ideas. And then I generally ignore the holes, because this is my editorial space and therefore my choices. But it’s good to debate.
Tech is still very biased and narrow-minded, due to the identities and interests of the people who build and own the tools. BIPOC lives, women, erased history, and left-of-center (humanistic, justice-aligned, people-over-profit) analyses are not baked into the systems, and are often de-prioritized. I’d discovered this years ago when starting the research on the first Journey Journal. And in this post-DEI era, one has to be sleuth, advocate, jury, and publisher—against the rig of the machines.
The writing process for me involves paper and pen, cursor and screen, dictating into my phone, meditating and working through words in my sleep and dream state, and editing myself to the nth.
I run the final material through two rounds of copy editing and fact checks. Fact checking is where I am most prone to “listen to” the agents because I play one off the other. It’s kind of fascinating and exasperating at the same time. Sometimes I send a final draft back to the original one and it stands itself corrected. In a perfect world where even solopreneurs had research assistants, this process might be streamlined. But I lived in academia once, so I ain’t so sure of that, honestly!
We fight a lot, me and the machines, at this stage because even though I’ve told them time and time again never to interpolate facts, make up scenarios, or use the dreaded “AI slop” language (empty words like “quietly,” three or four word sentences, horrible triplets—there are writerly triplets and then there are the ridiculous ones, which I state like this: “The table. The basket. The rock. You know, friend, you know.”) and the use of words like “rooms” to depict “situations” and “contexts,” etc. Those words smell so bad. Make us all quite dumber.
If you see them creeping into my text, I was tired that day and I am after all, a human dancing with and sparring with machines. Tell me and I’ll stomp it out like a roach.
One exception may be social media captions, because I used to write long, thoughtful captions and the posts would get five views. So I let the machines tell me how to play this silly algorithm game. It’s the lowest-hanging fruit. And it seems few people read captions anyway; are they just there to replace keywords because hashtags don’t function anymore? It’s a lot, and the bros controlling social platforms have deprioritized my topics for seven years, so I can’t concern myself too much. I recall when I thought Facebook ads were going to help and the only audience category selectable with the word “sisterhood” was people who were fans of the book The Sisterhood of the Traveling Pants. I wish this was a joke, but it’s true.
Comments though are always from my thumb.
The things I most rely on Claude for are coding, fixing website pages and architecture, building forms, quizzes, and logic, and creating databases and systems to keep all my materials organized. It also culls all my vast collections of written material and coursework so I become my own closed loop little learning machine.
This is not all free. I do need to sell some of my materials and generate income to keep this enterprise afloat. And a service I (formerly) used built “sales” pages throughout the site and across my offerings, many of which sounded pithy and pushy. I’ve tried to eliminate as many of these as possible.
Generally readers know my voice. I use big words, complicated construction, long long sentences… and I also insert short snippy asides with lots of slang and fourth-wall interjections. I hope the combination of this, my own opinionatin’, and the research and structural support I get from the agents all combine to provide an illuminating, entertaining, and helpful experience for you as the reader and learner.
I keep evolving and growing too, in this landscape. Thanks for coming along!
PS. I also use Claude and Google Translate to translate my writing into other languages. However, I only translate into languages I speak and read well enough to quality check these myself.