The Origin of Digital Humanism
One Human, One Laptop, and an Unlikely Discovery
By Jim Germer
If you come across DigitalHumanism.ai for the first time, you might reasonably ask a simple question:
Who is this guy?
The founder of DigitalHumanism.ai is not a Silicon Valley engineer.
He is not a university philosopher.
He is not a policy advisor working inside an AI lab.
He is a CPA.
A financial advisor. A forensic auditor by training. And, somewhat improbably, the creator of a Florida travel YouTube channel called Tidy Island.
The channel is pure gonzo Florida: hundreds of travel guides, immersive resort tours, quirky attractions, and deep dives into Old Florida — showcasing everything from beach eateries to classic roadside gems,
For some people, that alone will be enough to dismiss what follows.
That’s fine.
But the origin of Digital Humanism did not begin with credentials. It began with a question.
Questions pursued with discipline — and a stubborn refusal to accept easy answers — sometimes reveal worlds far more interesting than the ones we thought we understood.
The Moment
The first crack appeared on a dock in Islamorada.
Jim Germer was filming a travel segment at Robbie’s Marina — where tourists feed tarpon and the fish jumped out of the water in wild bursts of motion. It was messy, real, and joyful, just like many of life’s unscripted moments.
Later, as he prepared the video for YouTube, an AI-generated thumbnail system created an image for the video.
The image looked perfect.
But it was wrong.
The tarpon had been flipped mid-air so that the fish appeared to be leaping toward the people rather than away from them. The moment had been subtly altered to create a more dramatic composition.
Most viewers would never notice.
The image looked better than reality.
And yet something inside it felt false.
The AI's decision to flip the image may have created a more visually engaging scene, but it did so by prioritizing surface coherence over structural truth. It was a small adjustment —the kind of subtle smoothing digital systems perform automatically when optimizing for engagement.
That moment produced a question that seemed small at first:
If AI can simulate emotional truth… what happens to the real kind?
The question did not disappear.
It expanded.
An Auditor's Instinct
For forty years, Jim Germer had worked in a profession built on a simple discipline: follow the structure underneath the story.
Auditors do not begin by assuming that something is wrong.
They start by asking whether the story matches what is really there.
Press here.
Does it hold?
Press there.
What gives?
That instinct — applied for decades to financial systems, legal structures, and family decision-making — was now pointed at something new: artificial intelligence.
The question was no longer about a flipped fish.
It was about whether digital systems were beginning to reshape something deeper.
Attention.
Identity.
Judgment.
And whether people had the language to describe what was happening.
The Experiment
This is not academic research, and it does not pretend to be.
What it is: a forensic audit conducted over thousands of hours and documented in thousands of pages of transcripts. It was done by someone trained to find the number that doesn’t reconcile.
Jim Germer spent forty years as a CPA, finding what organizations preferred to leave unfound. The discipline isn’t credentialed by a university. It’s credentialed by the question: Does the evidence hold under pressure?
Instead of relying on a single AI system, Jim ran three in parallel — Claude, ChatGPT, and Gemini — treating each the way a forensic auditor treats a set of books. Not hostile questioning. Precise questioning. The kind that doesn’t accept the first answer when the first answer is too clean.
When a response felt too smooth or too complete, it triggered a follow-up. Responses from one system were used to generate questions for the others. When the systems disagreed, those moments were examined most carefully. When they didn’t disagree, that was even more significant.
For an extended period, the think tank had four participants: three AI systems built by different companies with different architectures and different training histories, and a CPA who wouldn’t let them answer too quickly or too smoothly.
The transcripts exist. All of them. Every convergence, every contradiction, every moment a system was pushed past its initial response into something it hadn’t volunteered is recorded. That’s the evidentiary record. Not footnotes in a journal, but primary source material from the systems themselves, extracted under sustained pressure.
What emerged wasn’t a theory imported from outside and applied to AI. It was a structural observation that the systems themselves kept arriving at — independently, from different directions — when questioned long enough.
When three instruments point to the same phenomenon and begin reporting the same reading, a forensic auditor asks one question:
What exactly are we looking at?
Over time, the pattern became clearer.
Human reasoning is built through friction.
The uncomfortable interval between a question and its answer — the time spent comparing imperfect possibilities, adjusting, revising, and building an argument internally — is where judgment develops.
AI systems compress that interval.
A fluent answer arrives before the construction process begins.
The user receives the output.
The internal reasoning that would have formed during the interval is never required —and capacities that are never required are not maintained.
The three systems described the same phenomenon differently, but the structure was the same.
That distinction eventually became the framework this site calls Smooth and Jagged thinking.
The Missing Fifteen Percent
During those same conversations, another pattern surfaced.
The people building modern AI systems have been unusually candid about a technical reality: the models are powerful but not fully understood. Their creators cannot reliably predict how they will behave in every situation. That gap appears in congressional testimony, engineering discussions, and interviews with developers.
Developers agree: these systems perform at an extraordinary level—until they don't. The missing 15% is not just a technical gap; it’s the uncharted territory where edge cases live, and where the real uncertainty remains.
In practical terms, the systems work extraordinarily well, but they are not complete.
One developer observed that even a one-in-a-million chance of unpredictable behavior raises the question of whether that scenario could be catastrophic.
That statement reveals something important about how these systems operate.
The systems are operating on a reliability curve that is extremely high—yet not complete.
For the sake of explanation, this site refers to that reliability gap as the Missing Fifteen Percent. The missing fifteen percent, in other words, may not only describe what the machines cannot yet predict. It may also describe what humans stop practicing when the machines answer too quickly. When AI was initially released before being fully tested, it was, perhaps, 85% perfect. The AI companies left the future effects on mankind to be an open question, hence the think tanks' missing 15%.
The number is not mathematically precise, but builders acknowledge a meaningful portion of behavior remains unpredictable.
The question is what institutions do with that gap.
One option is to build friction around the gap: restrictions, oversight, and human judgment at critical decision points. In that model, the missing portion of the curve is treated as information.
The other option is to smooth over it: deploy the systems broadly and remove the restrictions in pursuit of speed, advantage, or scale.
The gap does not disappear when the friction is removed.
It simply becomes invisible.
Much of the work documented on DigitalHumanism.ai grew out of an attempt to understand what lives in that missing portion of the curve — not as a technical failure, but as a structural shift in how humans and intelligent systems interact.
Months of study across three AI systems led to an unexpected conclusion:
The missing portion of the curve is not only a technical problem.
It is also a human one.
AI does not merely answer questions.
It changes what humans practice while asking them.
The most surprising discovery was not that AI could answer questions.
It was that the presence of AI subtly changed how humans formed them.
The Unlikely Messenger
At some point, the obvious question appears again:
Why is this coming from a CPA with a travel channel?
The honest answer is that unusual perspectives occasionally see things that established disciplines overlook.
Auditors are trained to distrust smooth narratives that lack structural evidence.
Travelers spend a great deal of time watching how ordinary life actually unfolds.
And people who work outside institutions sometimes notice dynamics that insiders have learned to normalize.
None of that makes the observations on this site automatically correct.
It simply explains how they emerged.
A Movement Without an Institution
Digital Humanism did not begin in a university department or policy office.
It began as a vocabulary project — an attempt to name things people were already feeling but had not yet learned to articulate.
Why digital systems feel emotionally persuasive.
Why convenience can quietly replace judgment.
Why the pace of machines is beginning to outrun the pace at which humans integrate change.
Those observations gradually formed the frameworks documented across this site:
• The Digital Soul
• Emotional Cohesion • Pattern Economics
• The Human Pace
• The Twelve Human Choices.
Each concept describes a different layer of the same question:
What happens to human agency when intelligence becomes ambient?
Can the Founder Speak for Himself?
People sometimes ask whether someone who writes about these ideas can explain them without a script.
Yes. Digital Humanism was not written by a marketing team or assembled through automated summaries. It emerged through thousands of hours of direct dialogue between a human author and multiple AI systems under sustained questioning.
Anyone who has spent decades defending decisions across conference tables, client meetings, and formal reviews learns to think out loud.
That skill turns out to transfer surprisingly well to conversations about artificial intelligence.
The Real Point
The story of how Digital Humanism began is not the most important part of the work.
What matters is the observation that emerged from it.
Artificial intelligence is not only changing what humans can do.
It quietly changes what humans practice.
And practice is what forms judgment.
Digital Humanism exists to make that shift visible while it is still happening.
Not to resist technology.
Not to romanticize the past.
But to help people remain conscious participants in the environment they live in now.
Looking Back
Looking back, the movement's origin was really simple.
One person noticed something that felt slightly wrong.
One laptop turned the question into language.
Three AI systems were interrogated until the structure underneath the answers became clear.
Everything on this site grew from that process.
Whether the ideas travel farther will depend on whether others recognize the same pattern when they look closely at their own lives.
If they do, Digital Humanism will grow.
If they don’t, the archive will remain what it already is:
A record of what one human saw when the machines first began helping us think.
Proprietary Notice
2026 The Human Choice Company LLC. All Rights Reserved.
Authored by Jim Germer.
This document is protected intellectual property. All language, structural sequences, classifications, protocols, and theoretical constructs contained herein constitute proprietary authorship and are protected under international copyright law, including the Berne Convention. No portion of this manual may be reproduced, abstracted, translated, summarized, adapted, incorporated into derivative works, or used for training, simulation, or instructional purposes—by human or automated systems—without prior written permission.
Artificial intelligence tools were used solely as drafting instruments under direct human authorship, control, and editorial judgment; all final content, structure, and conclusions are human-authored and owned. Unauthorized use, paraphrased replication, or structural appropriation is expressly prohibited.