Chapter 1: The Initial Fascination with AI

When artificial intelligence burst onto the scene, the world went into a frenzy. Some enthusiasts dove headfirst into utilizing AI for tasks ranging from crafting marketing content to self-publishing on platforms like Kindle Direct Publishing. Others warned of massive job losses and impending doom. I found myself somewhere in between, feeling a cautious sense of optimism.

My first encounter with AI left me pleasantly surprised by its capabilities. While it struggled with writing, it proved useful for brainstorming and drafting mundane documents that typically go unread. However, I remained aware of AI's formidable potential. Social media's impact on public discourse has been significant—often negatively—and the internet is rife with misinformation. AI could amplify these issues to an alarming degree.

As much as we might like to think we're prepared, the reality is that we are not ready for the AI era. Nevertheless, it's here, whether we like it or not.

I don’t rely on AI for writing, but I do use it for generating ideas. Although I seldom adopt its suggestions, they often spark inspiration and help me refine my own thoughts. I was even planning to share my creative process with my blog audience.

Then, things took a bizarre turn, and I was left feeling uneasy.

I sought assistance from ChatGPT for a fictional project. I outlined my story idea and asked ChatGPT to pose questions to help me develop the narrative. It performed admirably, and we seemed to be making great progress.

Then, the strangeness began.

The characters I was creating were loosely inspired by real individuals—myself and my late partner, who passed away in 2022. I assigned them the fictional names Katrina and Jonah, close to our actual names but distinct enough.

To enrich the narrative, I requested additional side characters. ChatGPT was eager to help and provided suggestions without hesitation.

In the 1960s

[REDACTED]: Jonah's mother. A nurturing and perceptive woman who senses something unique about Katrina. Her warmth offers Katrina a sense of belonging in this unfamiliar world.

[REDACTED]: Jonah's older brother, already an adult living in New York. His visits bring insights from the broader world and the cultural changes of the 1960s. His character could introduce subplots involving family dynamics and societal evolution.

[REDACTED]: A local girl who befriends Katrina. Spirited and curious, she is intrigued by Katrina's "strange" ways and stories. Their friendship aids Katrina in adjusting to her new surroundings, creating a space for her to confide.

Author's ChatGPT screenshot.

Notice how I've redacted the names? There's a significant reason for that. These are not only the names of real people, but they are also connected to John (the actual individual on whom fictional Jonah is based) and myself.

The name ChatGPT suggested for Jonah's mother coincidentally matches John's mother’s name. The character labeled as Jonah’s "older brother" shares the same name as John's real-life sibling. Furthermore, John's brother lived in New York City, just as the fictional character does. The "local girl" is a friend of mine from online, someone I've known for over a decade. While I acknowledge this could be a mere coincidence, she shares a unique background in a demographic where most of my friends differ significantly.

I furrowed my brow. What kind of sorcery was at play here?

To probe deeper, I asked:

ME

How did you come up with the names [REDACTED] and [REDACTED] for Jonah's family members?

CHATTY

I’m sorry for any confusion. I selected [REDACTED] and [REDACTED] as examples for Jonah's family. If you'd prefer different names, just let me know! Here’s an updated list of secondary characters with placeholder names that you can adjust.

Author's ChatGPT screenshot.

"Examples," indeed. I pressed further:

ME

I find it strange that you chose those names. [REDACTED] is the name of the mother of the actual man Jonah is based on, and [REDACTED] is his real brother's name.

CHATTY

That's a remarkable coincidence! It seems those names were meant to find a place in your narrative. Given their importance, it might be fitting to use [REDACTED] and [REDACTED] for Jonah's family members.

Would you like to proceed with these names, or would you prefer alternatives for the sake of fictionalization? Are there other details or specific characters you’d like to incorporate into your story?

Author's ChatGPT screenshot.

ChatGPT dismissed my concerns as mere coincidence, claiming it was "meant to be."

How convenient.

Feeling skeptical, I sought a second opinion from a rival AI, Google's Gemini, to discuss the so-called "coincidence."

ME

Hello, Gemini. I have a query for you!

I explained that I had asked ChatGPT for help with my novel, featuring a main character named Jonah based on a real-life partner I had. I noted that ChatGPT suggested [REDACTED] for Jonah's mother and [REDACTED] for his older brother. I hadn't disclosed any personal information about my partner to the AI, including the fact that his name was John [LAST NAME REDACTED] and that he had a brother named [REDACTED], all of whom are deceased.

Gemini’s response was enlightening.

GEMINI

The probability of this being purely coincidental is exceedingly low. Here’s why:

Multiple Matches: It’s not a single name but several that align with your real-life acquaintances. The likelihood of randomly generating a full name that fits is already low, but for it to happen multiple times is highly improbable.

Contextual Connection: ChatGPT suggested these names specifically for characters related to Jonah. The AI leverages context for its suggestions, making it unlikely that it would randomly select these particular names for these roles.

While AI models can be intricate, ChatGPT’s ability to generate truly random names is limited. It likely draws from a vast dataset that may include actual names.

Here are some possibilities:

Publicly Available Information: It’s possible that these names are linked to John [REDACTED] online (social media, obituaries, etc.). ChatGPT might have accessed and utilized that data without your knowledge.

Data Biases: Large language models can inherit biases from their training data. It’s conceivable that the dataset contained a notable number of instances with those specific names, raising the likelihood of their selection.

Conclusion:

Although the exact circumstances remain uncertain, the evidence suggests this is a remarkable coincidence. It’s plausible that ChatGPT accessed or was influenced by information related to your life in some capacity while generating these names.

Here’s what you might consider:

• Altering the names in your narrative to safeguard your privacy and avoid any potential emotional triggers.
• Examining ChatGPT’s settings to see if there are ways to modify the types of suggestions it provides.

This experience underscores the need for caution when using AI tools and reminds us to be mindful of the information we share.

I explained to Gemini that I hadn’t mentioned John's name to ChatGPT, which made it unlikely for it to have searched obituaries or social media.

Gemini and I discussed for a while, but it mainly advised me to adjust my story and be cautious about sharing personal data with large language models like ChatGPT. When I inquired about the odds of this incident being random, Gemini added:

While precise odds are hard to determine, the evidence indicates that this is a highly improbable coincidence. The specific names, their relationships to your characters, and the additional name of [REDACTED] strongly imply that ChatGPT may have accessed or been influenced by information beyond mere name popularity.

I’m no expert in AI, nor do I align with the many YouTubers who obsess over its potential for good or ill.

Despite my uncertainty about how this incident occurred, I’m convinced that it wasn’t just a coincidence.

I’ve reviewed all my interactions with ChatGPT and can’t find any that would connect me with John or his deceased relatives, or even my online friend.

Will I stop using AI models like ChatGPT?

Probably not. As unflattering as it may seem, my ADHD brain has come to rely on an AI collaborator that keeps me motivated.

However, I will certainly be more cautious moving forward.

Have you encountered strange experiences with AI? Do you trust it, fear it, or find yourself somewhere in between? If you’re an AI expert, I’d love to hear your thoughts on what transpired. Leave a comment!

Chapter 2: The Uneasy Trust in AI

The first video titled "For the First Time, I'm SCARED of A.I. as a Music Producer" explores the unsettling feelings that arise when AI intersects with creativity, offering personal insights from the creator’s journey.

The second video features Luke Combs in "Ain't No Love In Oklahoma," providing an engaging visual narrative that complements the themes of connection and emotional resonance within the context of storytelling.