I 10x'd Myself. It Cost Me the Boring Parts.
I ran the experiment on myself first.
By Mark Andrew Palmer · · 22 min read
I automated the boring half of my business and ran the experiment on my own company first. What changed, what I will never hand over, and what it bought.
Monday morning the brief was already written, and I had nothing to do with it.
It ran at six, read everything that happened across every company I own over the previous week, and put the whole thing on one screen. What moved. What stalled. What needs love. What is thriving. What is kicking ass and taking names. Three decisions it wanted from me, each one with the options laid out and a recommendation it was willing to defend. By the time I had coffee in my hand, the hard thinking about my own week had been done by something named Athena.
She is not a person. She is an agent I built, and I named her Athena because if you are going to bring on staff that never sleeps, never calls in and never needs a parking spot, you may as well give them names with some weight behind them.
There are a lot of them now. Argus works nights, watching the money and the inboxes and everything that moves while I am asleep, and in her first week she turned up three software subscriptions quietly billing me every month that nobody here had noticed, including me, and I am the man who signed up for them. Metis does research. Calliope drafts and never publishes a word on her own. Hermes handles correspondence, Peitho keeps the pipeline, Themis holds the line on anything that needs a human, Clio writes the case studies, Nemesis watches the competition, Hephaestus builds, and Mnemosyne holds every decision I have ever made and talks back when I ask her about it. There are more than that, and the list keeps growing, and I am not going to stand here reading you the whole roster like a man showing you his tackle box.
I am German and Irish, so by rights they should all be named Klaus and Siobhan and there should be considerably more drinking involved. I went Greek anyway. Partly because my entire company is built on the idea that myth is infrastructure and not decoration. Partly because the Greeks handed us democracy, geometry, tragedy and yogurt, which is a hell of a run for one peninsula and buys them permanent naming rights. And partly because "Athena wrote the brief" sounds like a company, while "Assistant Bot 4" sounds like a man losing an argument with his own software.
Names matter. That is not a cute observation, that is my actual profession. What you call a thing decides how everyone treats it, yourself included, and I have watched more than one good company go down because nobody in the building could say out loud what it was.
But the names are not the point and neither is the roster. Automation never arrives as a thunderclap. It arrives as a line item you forgot about, caught at two in the morning by something that never gets bored of looking. The point is not the staff. The point is what happened to my week.
I am 56 and I am having the best creative run of my life. I am aware the internet is currently wall to wall with people announcing their own renaissance, usually about ten seconds before they sell you a course about it. So let me do this the old way instead, with specifics, and with the company I keep.
Bicycle Day
On the afternoon of April 19, 1943, a Swiss chemist named Albert Hofmann sat down in his lab in Basel and took 250 micrograms of a compound he had cooked up himself five years earlier and shelved as useless. He thought he was taking a cautious amount. He was taking roughly ten times a noticeable dose. His notes say he took it at 4:20 in the afternoon, which is the sort of detail no writer would dare invent.
In case the picture has not fully arrived: the compound was LSD.
Lysergic acid diethylamide. Hofmann made it. Not a guru, not a poet, not a barefoot prophet on a hillside in a poncho. A career pharmaceutical chemist in a necktie, hunting a circulation stimulant for a Swiss drug company, who accidentally assembled the most notorious molecule of the twentieth century, filed it under useless, and then five years later went back and drank it to find out what it did.
The Beatles wrote it a song. Lucy in the Sky with Diamonds. Lennon swore to his grave the title came off a drawing his four year old brought home from school, and McCartney has since gone ahead and said the quiet part in public. Believe whichever version helps you sleep. Either way, every acid casualty, every psychiatric trial, every government that later lost its mind about it, every kid in a fur coat at a festival in August, all of it runs back to one man in a lab coat in Basel with a hunch and no volunteers.
Forty minutes in he wrote "dizziness, feeling of anxiety, visual distortions" in a hand that was coming apart, then asked his assistant to get him home. Wartime Switzerland, no private cars on the road, so they went by bicycle. That ride is now a holiday.
Here is the part people skip. Hofmann had already had the accident. Three days earlier a trace got into him through his fingertips and gave him two hours of what he called an uninterrupted stream of fantastic pictures. He had his result. He could have written the paper.
He took the full dose anyway, because a hunch is not a finding, and the only human being you can ethically hand an unknown compound to is the one who made it.
A hunch is not a finding. Somebody has to take the dose.
Eighty years later that molecule's grandchildren are being microdosed all over Northern California by product managers who take a tenth of a threshold dose on a Tuesday, feel approximately nothing, and then explain their protocol to you at a dinner party. Hofmann took ten times a full dose on his first deliberate attempt, alone, in a lab coat, with no idea what was coming, and then got on a bicycle. One of those men is running an experiment. The other one has a supplement routine and a newsletter.
There is a whole guild of the first kind and they are my favorite people in history. Werner Forssmann threaded a catheter through a vein in his own arm into his own heart in 1929, walked himself downstairs to the X-ray department to photograph the proof, and got fired for it. Nobel Prize, twenty-seven years later.
And my personal patron saint, Barry Marshall. July 1984, Perth, Australia. Marshall was 32 and convinced that stomach ulcers were caused by a bacterium, while every gastroenterologist alive knew for a fact that ulcers came from stress and that nothing could live in stomach acid. He said later that telling them otherwise was, in their ears, like saying the earth is flat. He could not test it on patients, because you cannot ethically hand a healthy person an ulcer. So one Tuesday morning he had his technician scrape two petri dishes into a beaker of beef broth, and he drank it like a shot. He did not tell his boss. He did not tell his wife. His wife found out the way everyone else did, which is to say she smelled it. (He did not ask her permission and she did not get a vote, so she got the results delivered straight to her face at close range for a week. Turnabout is fair play. But I digress.)
Within days his breath could strip paint, he was vomiting every morning, and a biopsy of his own stomach showed exactly the disease he predicted. The profession ignored him for another decade. Then, in 2005, they handed him the Nobel Prize in Medicine.
Go back far enough and it stops being science and starts being scripture. Odin, chief of the Norse gods, wanted the runes, which is to say he wanted knowledge. Nobody would hand them over, so he hung himself on the world tree for nine nights with a spear in his side, sacrificed to himself, myself to myself, and came down holding them. He also traded an eye for one drink from the well of wisdom. He did not send an intern.
Every one of these men, and the god hanging from the tree with a spear in his side, had the same two options you have got sitting in front of you right now. Wait for somebody else to run it and read about it later, or run it on yourself and know by the weekend.
And if a figure pierced in the side and hanging from a tree for the good of everybody sounds like a story you have heard somewhere else, you are not the first to notice, and I am not going to be the one to open that door on a Friday.
The boring half
Which brings me to the dose I have been taking all year.
Every business has two halves. There is the half that requires taste, judgment, nerve and a point of view. Then there is the other half: intake, logging, status, reporting, monitoring, chasing, filing, the weekly audit nobody reads, and the same problem being discovered three separate times by three people who never told each other.
For most of my career I paid for the second half with the first half's hours. Everybody does. You tell yourself it is the cost of doing business, which it is, right up until it stops being necessary and you keep paying out of muscle memory.
And people protect that half. They will fight you for it. A man who spends six hours a week assembling a report nobody opens has not got a task, he has built himself a job, and he will defend it in a meeting he also scheduled, using the word "process," while everyone nods because they have one too. Half the org charts in this country are a monument to work that could be done by a database and a decent night's sleep.
Alfred North Whitehead called this in 1911, in a book about mathematics of all things: Civilization advances by extending the number of important operations which we can perform without thinking about them. He was talking about notation, about the fact that you can multiply two numbers today without being a genius, because somebody invented a system that does the genius part for you. Same principle, every time. Every leap forward is somebody taking work that used to eat a person's entire attention and pushing it down into the machinery where it belongs.
Nobody ever built a reputation on the part of the job a machine can do.
So that half is gone. Not offshored, not dumped on a junior who quietly resents me for it. Gone into a system that files itself, watches itself, reports on itself, and comes and gets me when something needs a human with a spine.
Here is what surprised me, and it is the actual reason I am writing this instead of just enjoying it in private. I did not get free time. I got appetite. Every hour that came back went straight into the work with my fingerprints on it: the ideas, the naming, the design, the argument with a client about whether the logo really needs to be friendlier. (It does not, Deborah. It needs to be memorable. Friendly is what people call a logo when they cannot defend it.)
An Englishman named William Stanley Jevons worked this out in 1865, and nobody listened to him either. Watt's steam engine had just made coal power dramatically more efficient. Every engine now did more work on less coal, so every sensible person in England concluded the country would burn less of it. England burned vastly more. Because once power got cheap, they put engines in everything: mills, ships, mines, trains, whole industries that were never worth building at the old price. Efficiency did not reduce consumption. It detonated it.
That is precisely what happened to me. The work did not politely shrink to fit the free time. The appetite expanded to fill the new capacity, and it expanded fast, which is why I am building things this year I would not have said out loud last year.
Let it mow the lawn
Now let me tell you what I actually want out of all this, because the entire public conversation has it backwards and it has begun to irritate me.
Which, in fairness, is not a high bar this week. I am already irritated by your Birkenstocks, by the fact that you chew with your mouth open, and by whatever committee decided restaurants should be loud on purpose. This one, though, I can defend in front of a jury.
The panic is always the same shape. The machines are coming for the painting, the song, the novel, the poem. That is the part people say at dinner with their hand on their chest. And I understand the instinct, because those are the things we decided were sacred.
If I had my way, AI would clean my house. It would mow my lawn, fold the fitted sheet, and wipe my ass. It would sit on hold with the insurance company for fifty five minutes listening to a saxophone solo recorded in 1988. It would do my expenses, find out why the invoice bounced, and attend the meeting where three grown adults take turns reading an email that everyone already received.
Leave me the writing. Leave me the art. Those are not my chores. Those are the reason I get up, or, more accurately, the reason I am still up.
There is something profoundly stupid about a species that automated the harp and kept the dishes. And the laundry. And the hold music, the expense report, the permit renewal, the second follow up email to a man who is never going to answer, and the meeting that could have been a text message. We got the allocation exactly backwards, and a good chunk of this industry is sprinting to make it worse, building tools that write your wedding toast while you stand there sorting socks. Congratulations. You have outsourced the one thing in the room that was yours and kept everything a machine was born to do.
Nobody's soul lives in their expense report.
I do not want a machine that makes art. I want a machine that buys me the hours to make more of it. Those are not the same product, and the difference between them is the whole argument.
And here is the part that is not a joke, so I will say it plainly and then get back to being difficult.
I am 56. If the genes hold and I make reasonable decisions about bourbon, which is not guaranteed and is arguably not the plan, I have maybe twenty five good working years left. For the first thirty of mine, my ideas moved faster than anything available to build them. That gap is where careers go to die. Ideas do not wait around politely. They rot, quietly, in a drawer, and you tell yourself you will get back to them. I have that drawer. So do you. Mine had some good ones in it.
This is the first year of my life that the tools have moved faster than I do.
At 26 that would have been a fun fact. At 56 it is something else entirely, and the word for it is not excitement. It is closer to relief.
The report card with the D left in
I do not sell anything I have not first done to my own companies while the paint was wet.
Agency AEO is the clean example. I built it to get businesses recommended by AI platforms, ran the entire playbook on it, then went off to do client work and forgot to check the scoreboard for a month. When I finally looked, it was number two in the country for "AEO agency" off a single page. One page. No blog, no content team, no podcast, outranking shops with a hundred employees.
Since then I have been running experiments on it in public at a pace that would be reckless if the stakes belonged to somebody else. There is a new article on there most days. One of them maps where AI answers actually come from, and the finding should reorganize your marketing budget: roughly twenty domains supply about half of all citations in AI answers, Reddit alone is around a tenth of everything, Wikipedia another seven percent. Your beautiful website is not where the machine is looking.
My favorite is the report card. I took the same AI visibility audit I sell to clients, ran it on my own company, graded it honestly, and published it with my name on top. Two A's, two B's, and a D.
I left the D in.
Nobody in my industry publishes their D. Which is exactly why I did, and also why you should squint at any agency whose case studies are a wall of triumphs. A lab that only reports its successes is not a lab. It is a showroom. Go read your current agency's case studies and count the failures. Take your time. I will wait.
Dose is the whole argument
Now the part where I disappoint both camps.
One camp says AI is coming for everything you love. The other says it is magic and you can fire your staff by the end of the month. The doom camp is working on a book. The magic camp is working on a course. Between them they have shipped exactly nothing you can log into, and both of them will be on a panel about it next month.
It is a tool. A nail gun is a tool. It does not replace the carpenter, it replaces one motion of the carpenter's elbow, and it will still put a nail through your hand the minute you get cocky with it.
The best research anybody has run on this comes out of Harvard, MIT and Wharton, who put 758 Boston Consulting Group consultants through it in 2023. I trust it because the results cut both ways. On work that sat inside what the authors named the "jagged technological frontier," consultants using AI finished roughly 12 percent more tasks, about 25 percent faster, at meaningfully better quality. On work that sat outside it, the same tool made them measurably worse, and here is the line that should stop you cold, more confident while being wrong.
The confidence comes free with the wrong answer. That is what makes it dangerous.
You already know this guy. Every industry has one. He is never uncertain, he is wrong about four things a quarter, and he has never once circled back. Now hand him a tool that renders his certainty in clean paragraphs with a bulleted summary. That is not a man with AI. That is a man with a bigger megaphone taped to the same mouth.
Which is a dosing problem, and dosing has a patron saint too. Mithridates, king of Pontus, spent his life terrified of being poisoned, so he took a little poison every day to build his tolerance. It worked. It worked so well that when Rome finally closed in and he tried to take his own life with poison, nothing happened. He had to ask a guard to do it with a sword. Housman put the whole moral in four words: Mithridates, he died old.
So my rule has not moved all year. It never touches the wheel. It does not send money. It does not publish a client's words without a person reading them first. It does not write brand strategy, because strategy is a judgment about meaning and it does not have any. It drafts, watches, files, remembers, and argues with me on request. Then a person with taste decides. Small doses, every day, deliberately.
And that rule is not sentiment. It is not me protecting my feelings about craft. Taste is the only asset in this business that cannot be borrowed, copied, subscribed to or generated, and the minute you hand it over you have nothing left to sell but speed. There are going to be a great many very fast, very cheap, very forgettable companies over the next few years. They will all sound like each other. That is not a prediction, that is arithmetic.
Nobody left to ask
Here is the change I did not see coming, and the one I would go back and hand to my younger self.
I was the top guy at every company I ever started. Founder, principal, name on the paperwork, final say. And I still ran things past people, every time, because I had hired a C suite and grown adults deserve a vote in the thing they are building with you. That is not weakness. That is how you keep people worth keeping. I did not need the approval and I asked anyway, because asking was the decent thing and because the people I hired were worth listening to.
There is nobody to ask now. No partners. No committee. No approval cycle, no steering group, no pleasant man explaining that the interesting part is phase two and the budget is phase never. No platform standing between a thought I have at ten at night and a working version of it at two in the morning.
Understand what that actually deletes. It is not the time. It is the tax on ambition. When every idea has to survive three meetings and a quote, you stop having the expensive ideas. You do not decide to stop. It happens quietly, the way a man stops ordering the thing he wants because he already knows what it costs, and after a few years he believes he never wanted it.
I am a designer. I have been one my entire adult life, and before that I was a kid in the West End who could see the finished thing before he could build any of it. Which brings me to the oldest wound in every studio on earth, and if you make anything at all you already know exactly what I am about to say.
Your imagination is not the constraint. It never was. Your hands are.
Every artist alive is capped at the precise width of their own technical ability. You see the thing whole, in color, finished, better than anything else in the room. Then you sit down and produce a version of it degraded by exactly how good you happen to be at the craft that particular Tuesday. The distance between what you saw and what you managed to make is the thing that wakes artists up at four in the morning, and no amount of talent closes it, because talent widens the vision at the same rate it improves the hands.
I went to school for sculpture. Metal. I know precisely what it is to carry a vision in your head and a weld that will not hold it.
T.S. Eliot wrote the whole problem in 1925 and did not need a paragraph: Between the idea and the reality falls the Shadow.
Leonardo died with notebooks full of machines that could not be built. Flying machines with no engine to put in them, because the engine would not exist for four centuries. He was not wrong. He was early and unequipped, which history files under the same heading and which every one of us has been at some point, usually on a smaller and less flattering scale.
The vision was never the problem. The hands were.
That gap has not closed. Let me be precise, because this is the part people oversell. It has narrowed. I still need the taste to know what is worth building, the judgment to kill the bad version, and an idea that deserves to exist. Nothing has automated any of that and nothing is about to.
But the distance between the picture in my head and a working version of it on the internet is now measured in hours. Not quarters. Not budget cycles. Hours.
So I build the thing the week I think of it. Not a mockup of the thing. The thing, working, live, with a login. In the last month: a brain for my whole company that I talk to out loud and that answers in a voice I designed for it. A dashboard that runs every property I own. A deck of fifteen brand archetypes with a site of its own. And the one I am most childishly happy about, websites you walk into instead of scroll down.
Those are demos. Experiments, unfinished on purpose, and that is the entire point. The experiment is the work now. I am not selling you a product out of that link, I am showing you what a website can be once nobody has to file a ticket to find out. Go stand in one rather than let me describe it, because describing a room is not the same as being in it.
None of it required me to become an engineer. It required me to stop accepting that the distance between an idea and a working version of it gets measured in quarters.
Kid in a candy store is the cliche everybody reaches for, and it is not quite right. It is closer to being handed the keys to the kitchen after twenty years of ordering off the menu and finding out you can cook.
I do not need permission anymore. Not from a partner, not from a committee, not from a developer's calendar, not from a platform's roadmap. My creativity does not need an excuse, a budget line or a sign off, and I am telling you this not as a victory lap but as a public service, because here is the part that should genuinely unsettle you:
Neither do you. You just have not tested it.
On the phrase "cracked the code"
People keep saying that one to me, and I want to handle it carefully, because it implies a trick, and there is no trick.
There is no prompt. There is no app. There is no twenty dollar a month subscription that does this. Anybody selling you one has sold you a hammer and called it a house. I have watched business owners buy four of those hammers this year and wonder why the roof is still open.
What there is, is a system. One place that holds everything all of my companies know, so nothing has to be remembered twice. Agents with real jobs, real names and hard limits. A written rule about what they are never permitted to touch. And underneath all of it, a hundred small experiments, most of which failed, a handful of which changed how an entire company runs. The report card with the D on it is one of those experiments. So is this essay, incidentally, and you are standing in the middle of it.
If you want the whole thing in one sentence: decide what only you can do, build the thing that does the rest, and be brutally honest about which list each task belongs on.
That last clause is where almost everybody fails, and not for lack of intelligence. People put things on the wrong list on purpose. The boring work is safe. Nobody has ever been humiliated by a spreadsheet. The real work, the work with your name on it and your taste exposed, is frightening, and a calendar packed with the safe half is the most socially acceptable hiding place ever invented.
I owe you the long version of that, with the actual architecture in it, and it is coming. Not tonight. Tonight I am still on the bicycle.
What the 10x actually bought
Not hours. The day is still the day, and I am not one of those men who claims to have found a secret ninth hour in the morning. Hell, if I am up in the morning, I am still up. Not freshly awakened.
What it bought is that the boring half runs whether I am in the room or not, so the half that requires me gets all of me. My work is better than it was a year ago. Not more prolific. Better, which is a harder claim and a checkable one, so go check it.
And it bought availability, which I am going to say plainly one time and then leave alone, because the whole point of this piece is that I am not pitching you.
Somewhere in this town there is a guy named Dale paying a monthly retainer for a report he does not read, written by somebody drowning in precisely the work I just described. That agency is not crooked. It is just spending its whole week on the half a machine should be doing, and charging Dale rent on the delay. Dale calls and waits four days, and has decided that this is what professionalism feels like. My phone gets answered every time. Sometimes by an agent I built for exactly that purpose, sometimes by me, and which one of us you get depends entirely on the caller ID. That is not rudeness. That is the system working precisely as designed. The machinery does the filing, I do the thinking, and I am selective about who I work with for reasons that have nothing to do with being busy.
You are a business owner. Drawing conclusions is the job.
I will not end this with a warning. I am tired of warnings and so are you.
Everything above was built by one man in a room in Louisville, using tools available to anybody with a laptop and enough stubbornness to sit there until they work. The barrier was never access. The barrier is that most people will read this, agree with it, feel a pleasant little rush of intention, and then go answer email. A few of them will forward it to somebody else, which is how a grown adult does nothing while appearing thoughtful. One or two will leave a comment about how AI lacks the human touch, typed on a device that finished the sentence for them.
Hofmann got on the bicycle. Marshall drank the beaker. Odin gave up the eye. Nobody in that lineage waited for the industry to come around, and every one of them was right years before the medal showed up.
I am not going to do the humble thing here, because you would smell that too. I see this stuff early. I always have, it is the one gift I got, and by itself it is worth absolutely nothing. Everybody on my street saw AI coming. Seeing it is free. The rare part, the only part that has ever paid anybody, is being the one who drinks the beaker on a Tuesday morning while the whole profession stands there explaining why you are wrong.
That part is available to you. Today. For the price of being early and slightly ridiculous in public.
Two hundred fifty micrograms. A beaker of beef broth. An eye at the bottom of a well. Not one of those men got a guarantee, and not one of them got the medal on time. Forssmann got fired. Marshall got laughed at for a decade. Hofmann got a molecule that outran him for the rest of his life and a great deal of blame for what strangers did with it. What each of them got first, before any of that, was the answer.
And a hell of a ride home.
Take the dose yourself. Keep the art.
- MP
Tags: Business, Agency Model, Agencies, Branding, Brand Strategy, Entrepreneurship, AI, Automation, Creativity, Agency AEO, Omythic, Mark Palmer