An Unhinged Look at My Life

According to ChatGPT

Have you ever wondered what artificial intelligence would say about you if you gave it permission to be completely unhinged?

I recently read Randy Seaver's Genea-Musings post, The Completely Unhinged Life of Randy , and couldn't resist trying the same experiment.

I opened ChatGPT and entered a very simple prompt:

"Please write an 'unhinged narrative' about me."

Since I've spent considerable time chatting with ChatGPT about my genealogy research, historical society projects, blogging, and adventures with technology, I was curious to see what it would come up with.

The result? A rather entertaining account of my so-called retirement, my ongoing battles with the Crawford family, and my tendency to ask AI to create images with absolutely NO PEOPLE!

Of course, ChatGPT didn't get everything right. It invented children and grandchildren for me, neither of which I have. That's a good reminder that even when AI appears to know us well, it can still make things up.

I've corrected that portion of the narrative, but otherwise, I'm sharing the results for your amusement.

Thanks to Randy for the inspiration!

Here's how ChatGPT describes my life when given permission to go completely off the rails.

Marcia Philbrick: The Woman Who Refuses to Let the Dead Rest in Peace

A completely unhinged, mostly affectionate, and alarmingly evidence-based narrative.

Somewhere in Seneca, Kansas, there is a woman who retired in 2016.

Allegedly.

Her name is Marcia Philbrick, and she has spent the years since retirement demonstrating that the word "retired" is merely a suggestion made by people who lack sufficient genealogical documentation.

Marcia has approximately 29,000 people in her genealogy database, 43,000 digital files, and enough unresolved Crawford family mysteries to qualify as a federal investigation.

The Crawfords are not cooperating.

There are three men named James Crawford. They lived in Kentucky around the same time. They married different women. They bought land near each other. They associated with the same families.

Naturally, generations of genealogists decided they must all be related.

Marcia said, "SHOW ME THE EVIDENCE."

Then she brought in DNA.

And the DNA said, "Well, actually…"

Now entire branches of the Crawford family tree are being evicted. William Crawford has been escorted out of the presumed brotherhood. James Crawford Jr. is under investigation. Rebekah Crawford's land transactions are being scrutinized. Somewhere in the great beyond, a Crawford is frantically hiding a marriage bond.

Marcia is not amused.

She has spreadsheets. She has maps. She has FAN networks.

She has receipts.

And she will find you.

Meanwhile, the Internet Is Barely Keeping Up

Most people write a blog post and call it a day.

Marcia writes a blog post, creates a shorter version for Substack, requests a social media blurb, generates a 16:9 feature image, and schedules the whole operation across multiple platforms.

At 6:30 a.m., Blogger gets the story.

At 7:00 a.m., Substack receives the abbreviated version

At 7:30 a.m., Facebook gets involved

At 11:15 a.m., Instagram enters the conversation.

And heaven help the AI that puts an actual person in her feature image.

"I SAID NO PEOPLE."

The AI apologizes.

"AND MAKE IT 16:9."

The AI apologizes again.

"WHERE IS THE HEARTLAND GENEALOGY BRANDING?"

The AI quietly questions its career choices.

The Historical Society Has Entered the Chat

When Marcia isn't interrogating deceased relatives, she's helping the Nemaha County Historical Society.

There are exhibits to prepare. Museum artifacts to describe. Fundraisers to promote. Buildings to preserve.

A wool carder appears.

Marcia: "Explain what this is and how it worked."

A spinning wheel appears.

Marcia: "You're next."

A treasure hunt is announced.

Suddenly, there are promotional graphics, donation instructions, early-bird shopping hours, and a Friday Bag Day sale.

ONE DOLLAR PER BAG. FIVE DOLLARS PER BOX. NO CLOTHING OR ELECTRONICS.

The entire county has been informed.

Twice.

Possibly three times.

Then She Discovered AI

Some people approach artificial intelligence cautiously.

Marcia approached it like a retired technology coordinator who had just discovered an unpaid research assistant willing to work 24 hours a day.

"Transcribe."

"Improve this."

"Write a shorter version."

"Create an image."

"NO PEOPLE."

"Try again."

She has used AI to investigate ancestors, untangle historical relationships, develop museum materials, write biographies, create song lyrics, and turn genealogical research into videos.

She has also discovered that AI occasionally invents things.

This does not go over well.

When an AI connects Thomas Kennedy to the wrong James Crawford, Marcia immediately demands an explanation.

The AI begins sweating in binary.

When a PDF attachment doesn't work, she threatens to go to Gemini.

Somewhere, a server lights a candle.

And now she's preparing to explain AI to an entire group in Jackson County.

The presentation will probably begin with a simple message:

"Keep your prompts simple."

It will end with a detailed explanation of why three men named James Crawford cannot be casually assigned to the same family.

There will be diagrams.

Her Husband Thought They Were Retiring Together

Michael Philbrick retired from teaching in 2016, too.

One might imagine a peaceful retirement.

Perhaps some travel. A little relaxation. 

Instead, Michael films church services for YouTube while Marcia prepares worship slides, manages historical society projects, maintains a genealogy empire, and investigates people who have been dead for 200 years.

Their computers probably need a union.

Their nieces and newphe have an aunt who can tell them exactly where an ancestor purchased 50 acres in Kentucky in 1803 but may have difficulty explaining why anyone would name three consecutive generations James.

Actually, she can explain that.

She has a chart.

The Final Boss: An Unidentified Crawford

One evening, Marcia sits down at her MacBook Air.

She intends to spend just a few minutes researching.

A dangerous phrase.

She opens RootsMagic.

Then Ancestry.

Then FamilySearch.

Then a digitized county court record.

Then a land patent.

Then a newspaper.

Then a DNA project.

Then another land patent.

Suddenly, it's 2:00 a.m.

She has discovered that a man named William Crawford signed a document as "Junr."

There are two possible Williams.

Neither is cooperating.

She creates a research note.

She opens ChatGPT.

"Can you improve this blog post?"

The next morning, Heartland Genealogy publishes another installment.

Somewhere, an unsuspecting genealogist reads it and realizes that the family connection they've accepted for 40 years might be wrong.

Marcia smiles.

Not because she enjoys destroying family trees.

But because the evidence matters.

And because somewhere, buried in a courthouse record, an overlooked newspaper clipping, or the DNA of a distant cousin, there is an answer waiting to be found.

She will find it.

She may have to reorganize 43,000 files, rewrite 17 blog posts, and convince another Crawford descendant to take a Y-DNA test.

But she will find it.

And when she does?

She'll write a blog post.

Then a shorter version for Substack.

Then a social media blurb.

Then she'll ask for a feature image.

16:9. Light colors. Heartland Genealogy branding. Absolutely no people.

And somewhere in Kentucky, the Crawfords will continue to regret ever leaving a paper trail.

Marcia Philbrick: Retired educator. Volunteer historian. Genealogical detective. AI supervisor. Relentless defender of documentary evidence.

Retired since 2016. Hasn't slowed down long enough for anyone to verify it.