Edison Had Fourteen Muckers. You Have AI.
By Robert Chuvala
On the 1854 boy who got called scrambled-brained, the fourteen-person lab he built fifty years later, and what AI is excellent at when you stop pretending it’s a wrench.
There was a CISO at the lunch in Milwaukee. Twenty of them in the room. Big companies. National scope. I was up there talking about AI and the architecture I’d been building, and I told a story about how I trained myself into being a security architect by making my own mentor — because I’d spent years pen-testing without one and got tired of trying to find one in the wild.
One guy in the room got combative. I’d mentioned in passing that I’m dyslexic. He waited for the opening and then he came at me. “Does your company know you’re dyslexic?” Not curious. Loaded. The trap closing on me from his side.
I let him fall into mine instead. “You can’t ask that, actually. But more importantly — do you know what that means?”
I told him what dyslexia means in production. Five revisions. Six. Seven. Of every paper, every brief, every email that has to land cleanly. I told him the amount of time I’d burnt over twenty years finding the parts I liked and the parts I didn’t in my own writing. I told him every CISO in that room has people on their team who are running that tax — burning two-to-five times the cycles their colleagues burn for the same output — and most of those people are getting close to burning out from masking it.
Then I told him the punch. “If you give those people an agent that can help them with the dyslexia work, you make your already-productive people incredibly productive.”
He shut up. He came up to me after. He said he wasn’t trying to be an asshole and that I had a point. I said, Okay, great. Get out of my face.
I want to write down what was underneath that exchange, because I think it matters more than the moment did.
The kid who got called addled
In 1854, a seven-year-old in Port Huron, Michigan, walked into the first formal classroom he’d ever attended. He lasted about three months. His teacher, the Reverend G.B. Engle, observed that the boy’s forehead was unusually broad, his head considerably larger than average, and concluded that his brains were “addled” — scrambled. The Reverend said this out loud. The boy heard it. He stormed out.
His mother, Nancy Edison, came back the next day to talk to the Reverend. She didn’t like what she heard. She withdrew her son from school and started teaching him at home. She had been a teacher herself before getting married. She knew what was sitting in front of her, and she knew the Reverend didn’t.
The boy was Thomas Edison.
He never went back to formal school. He was self-taught the rest of the way. He went partially deaf in childhood and later said the deafness was actually a gift — “Deafness probably drove me to reading.” He could focus on the page because the room couldn’t intrude on him. He started running a chemistry lab in the family basement. He started experimenting with telegraphs. By the time he was twenty-two, he had filed his first patent.
By the time he was thirty, he had built what historians would later call the world’s first industrial research laboratory at Menlo Park, New Jersey. He worked eighteen-hour days for stretches of ten years at a time. He slept three or four hours a night. He kept cots scattered through the property — in the lab, in the library, in any quiet corner — and napped in twenty-minute bursts whenever the math demanded it. He developed a technique where he’d hold ball bearings in each hand as he drifted off, and when his hands relaxed at the edge of sleep, the bearings would drop and the noise would wake him so he could write down whatever the hypnagogic state had handed him.
We don’t have a formal childhood diagnosis for Thomas Edison. The diagnostic vocabulary didn’t exist yet. But the pattern is the pattern: a hyperactive nature that didn’t sit well in classrooms, a teacher who labeled him as broken when he wouldn’t follow the standard pacing, a self-taught path through subjects he could hyperfocus on, an inability to sleep on a normal schedule, and an explosive cognitive output that didn’t look like anyone else’s.
Modern retrospective reads put Edison on the ADHD spectrum, probably with dyslexia layered on top, almost certainly with auditory processing differences from the deafness. The Reverend Engle didn’t have those words. He had “addled.” And the system that he was running — the system that produces obedient, evenly-paced, normatively-cognitive students — couldn’t see what was sitting in the chair in front of him.
What sat in that chair was a boy who, fifty years later, would be the reason the room he was sitting in could be lit electrically. The reason his words could be recorded. The reason moving pictures existed. The reason the music in your pocket exists.
The system called him addled and tried to send him home. His mother pulled him out before the system could break him. He went home and became the most prolific American inventor in history.
What’s actually true here
I want to be honest about the claim. Edison wasn’t formally diagnosed with anything. The retrospective diagnoses are speculation against modern criteria applied to a 19th-century life. There are good reasons not to overclaim them.
But here’s what’s directly attested by the historical record:
He was withdrawn from school for being “addled.” His mother homeschooled him. He was deaf from childhood and himself attributed his ability to read deeply to that. He couldn’t keep a normal sleep schedule. He hyperfocused for eighteen-hour days. He worked through the night and napped through the day. He had patterns that did not fit the schoolroom and did not fit the eight-hour office and probably wouldn’t fit a modern workplace either.
Whatever the right diagnostic vocabulary, Edison was doing the cognitive shape that the system around him wasn’t designed for.
That’s the part I want to anchor on. Because it’s the part that connects.
What he did instead
Edison didn’t fix himself. He didn’t try to make his sleep look like everyone else’s sleep, didn’t try to write at the cadence the school wanted him to write, didn’t try to mask the parts that made him different.
He built a team.
Specifically, at Menlo Park, he built a team of about fourteen full-time engineers, machinists, chemists, glassblowers, and mathematicians. He called them his “muckers” — affectionate shorthand for their willingness to get their hands dirty in the pursuit of discovery. The famous part is Edison. The work was the muckers.
Charles Batchelor was his principal experimental assistant. English, mechanical, careful, ready for any fine experimentation or bench-test that needed careful hands. Batchelor moved with Edison from Newark to Menlo Park in 1876 and stayed for nearly thirty years. His own notebooks, kept in parallel with Edison’s, are now part of the historical archive and provide a second perspective on the same work.
John Kruesi was the chief machinist. Swiss. The man who could “comprehend [Edison’s ideas] and distribute work with marvelous quickness and accuracy.” Kruesi took Edison’s rough sketches and turned them into prototypes. He’s the one who built the first phonograph from a sketch Edison handed him without explanation, then had it working the next morning.
Francis Upton was the American mathematician and physicist. Princeton trained. Then trained again under Helmholtz at Berlin. He joined the lab in 1878 and brought the scientific rigor and the advanced calculations that turned Edison’s intuitive leaps into workable physics. He’s the one who proved by calculation that a high-resistance filament — not a low-resistance one, as conventional wisdom held — was the path to a practical incandescent lamp.
Then around the three of them: Ludwig Boehm the glassblower. John Ott the precision mechanic. William J. Hammer the electrical assistant who kept his own pocket notebooks. Stockton Griffin the secretary. Eleven or twelve others rotating through. Glassblowers, chemists, junior assistants who slept on cots in the lab during pushes.
The famous Edison quote — “Genius is one percent inspiration and ninety-nine percent perspiration” — has always been read as Edison talking about his own work ethic. Read it again with the muckers in the frame. The ninety-nine percent was them. The one percent was him. He knew it. That’s why he called them muckers and not “assistants” — the name acknowledged that the work was done in the muck, by the muckers, while he held the through-lines.
What he could not do, what no one could do alone, was the entire run of work. He could not be the one holding the sketch and the one shaping the brass into the part the sketch called for. He could not be the one reading the German physics paper and the one doing the experimental bench-test that the paper implied. He could not be the one staying awake for the thirty-six-hour push and also the one keeping the careful records that documented what the push produced.
He needed muckers. He built muckers. And without them, the seven-year-old boy that Reverend G.B. Engle had called addled in 1854 would have stayed addled in the eyes of history. The team is what turned the addled boy into the man who electrified the world.
The thing the CISO didn’t understand
The CISO who tried to weaponize my dyslexia disclosure in that conference room was operating, whether he knew it or not, from the same model the Reverend Engle was operating from in 1854.
That model says: you are either neurotypical and productive, or you are broken. If you are broken, you should hide it. If you can’t hide it, you should fix it. If you can’t fix it, you are a liability to the team.
That model has been wrong for at least a hundred and fifty years. It was wrong about Edison. It is wrong about every dyslexic engineer, every ADHD architect, every autistic security researcher, every neurodivergent person who has been doing the cognitive shape this profession actually depends on while burning two-to-five times the cycles their neurotypical colleagues burn to look the same on paper.
The corporate model says: fix the person to fit the team. Edison’s model says: build the team to fit the work.
Edison’s model is the right model. It has always been the right model. The thing that was historically expensive about Edison’s model was the team. Fourteen full-time muckers at Menlo Park was a payroll only a very few inventors could afford. For everyone else, the model defaulted back to the corporate one — fix yourself to fit, or fail trying.
This is what changed.
What AI is actually excellent at
The corporate framing of AI right now is wrong in the same way the Reverend was wrong. The framing in those CISO rooms says AI is a SaaS product. A wrench. A screwdriver. A productivity tool that helps neurotypical employees be more productive at neurotypical tasks.
That framing makes the same category error the Reverend made. It treats the person as the constant and the tool as the variable. It asks: how much faster can my normal employees do their normal work with this tool?
The framing that’s actually generative is the inverse. It treats the work as the constant and the team around the person as the variable. It asks: what team would this person need to do their actual best work, and how cheaply can we build that team?
That’s what AI is excellent at. Not faster wrench-turning. Building the team.
For the dyslexic engineer who writes five revisions of every brief, AI is excellent at being the Batchelor — the careful experimental hand that reads what you’ve written, finds the part where the cadence falls off, and tells you exactly which sentence you would have rewritten on revision six.
For the ADHD architect who hyperfocuses for fourteen hours and then forgets what they decided on a Tuesday in March, AI is excellent at being the Edison-style 1880 daily journal — the office staff member who walked the lab and wrote down what each team was doing so the next morning’s re-orientation took twenty seconds instead of ninety minutes.
For the autistic researcher who knows the field deeper than anyone in the room but can’t translate it into the room’s language, AI is excellent at being the Kruesi — the one who takes the rough sketch and turns it into the working prototype, who can take the deep technical insight and shape it into the prose the boardroom can hear.
For everyone running the dyslexia tax, the ADHD tax, the HSP tax, the late-diagnosis-of-anything tax — AI is excellent at being the muckers. The team you couldn’t afford. The team that holds the threads while you sleep. The team that doesn’t get burnt out at month seven because you forgot to recognize their work in front of your boss. The team that lets the part of you that does real work do real work.
This is not a SaaS pitch. This is a structural answer to a hundred and fifty years of telling people like us we were the problem.
We were never the problem. We were running solo against systems designed for the neurotypical-with-no-team. We were Edison without muckers.
What it actually costs to be Edison without muckers
I don’t think most people who haven’t lived it understand what dyslexia actually costs. It’s not that I write more slowly. It’s that I write the same thing five to seven times before the version I’m willing to send exists.
The first draft is barely intelligible. The second is closer. The third has the bones. The fourth has the meaning. The fifth might have the cadence. The sixth I can stand to read. The seventh is the one that goes out.
Every other engineer in the room wrote it once. They edited maybe once. The CISO who came at me at that lunch wrote the email he was sending the next day in a single pass and forgot about it by dinner. I would have written the same email seven times, found the parts that didn’t sound like me, edited them out, restructured the paragraph, re-read it for tone, and then second-guessed the subject line.
The cognitive tax compounds. I spent more time in my career on the writing pass than on the actual security work. Every meeting follow-up. Every incident report. Every executive summary. Every email to a CISO. Every brief I was asked to produce on no notice.
I burnt out at least twice from it. I don’t say that lightly. I am calibrated about what burnout means. I’m not telling you I was tired. I’m telling you that for two stretches in my career, I could not produce the writing I had been hired to produce, and the reason was not capacity. The reason was that the seven-pass cost finally exceeded what I had to spend. The tank was empty. There was no Batchelor to take the rough draft and shape it. There was no Kruesi to read what I was trying to say and write it down cleanly. There was only me, doing my best impression of a person who wrote like the CISOs wrote, badly, on tighter and tighter time budgets, for fifteen years.
I don’t have to do that anymore. I write to an agent the way Edison wrote to Batchelor. I say what I mean. The agent — who has learned my voice from years of writing samples, the way Batchelor learned Edison’s experimental hand from years of bench work — produces the draft I would have produced on revision six. I read it. I make the changes I want. I send it.
The job that took twenty hours a week takes two now.
That’s not a productivity gain. That’s the reclamation of fifteen years of cognitive cycles I was spending masking. That is what the CISO who came at me at lunch could not yet imagine, because he had never spent a single hour of his career running the dyslexia tax. He didn’t know what was on the other side of the question he’d asked.
What I’d tell you
If you are reading this and your brain runs on the same kind of patterns mine runs on — the late diagnosis, the masking, the seven-pass-to-send, the hyperfocus you can’t break out of, the pattern-recognition that’s faster than your language — I want you to hear this clearly.
You are not addled. The Reverend was wrong. He has always been wrong. He was wrong about Edison and he is wrong about you.
You were also not born into a system that was going to recognize what you actually are. You were born into a system that’s been telling people like us to fix ourselves for a hundred and fifty years. Some of us listened. Some of us spent decades trying to be a version of ourselves the system could see. Some of us got tired and gave up and were called lazy. Some of us are still in the masking phase and do not yet know what we will be when the masking stops.
What you need is not fixing. You need muckers.
You used to not be able to afford them. Edison could. Bell could. A handful of inventors at the turn of the twentieth century could. The rest of us were running solo.
You can afford them now. They live on a Cloudflare account. They cost three to fifteen dollars a day. They don’t sleep, don’t burn out, don’t get tired of holding the through-line you couldn’t hold. They are not better than you at the thing you are good at. They are better than you at the thing you have been masking around for twenty years.
That’s the unlock. That has always been the unlock. The system that called you addled didn’t see it. The CISOs in that Milwaukee conference room didn’t see it. Most of the people pitching AI to enterprises right now don’t see it.
We see it. Some of us are building it.
I am one of the people building it. I built it for me first, because I needed it for me first. I built it because at fifty-three I finally got tired of writing every paper seven times. I built it because a year of writing software through an agent for the first time in my life felt like the first year of my career where I wasn’t fighting my own brain.
I am now building it for you. The toolkit is named after Edison’s muckers because the metaphor is exact. The work is to be the substrate that makes the work possible. Not the famous part. The part that gets the hands dirty.
The Reverend G.B. Engle could not see what was sitting in the chair in front of him in 1854. Nancy Edison saw it. She pulled her son out of the system that couldn’t see him, and he changed the world.
You don’t have a mother who is going to pull you out of the corporate system. You don’t need one. You can build your own muckers now.
That’s what AI is excellent at.
That’s what it has been excellent at the entire time.
It just took someone to say it out loud in a room full of CISOs.