Execution over everything

The distance between an idea and an outcome is closing. Who are you becoming?

Rain, Steam, and Speed - The Great Western Railway by Joseph Mallord William Turner

Some days I finish work and cannot utter a word. Not because nothing happened, but because too much did. My brain spent eight hours switching between strategy, branding, code, writing, management, and decisions, then circling back before any one context had fully cooled.

I recently told a friend, “Yeah, I have twelve subagents cleaning up all the slop code the previous models made. But look at the results. This search result is thirty percent faster now.” He had no use for this information, though he knows I’m building an archive of public-domain artworks.

Then I heard myself and laughed. What world are we living in? Maybe the bigger question is: Who am I becoming?

I see people hold their laptops up to their mouths like water fountains and whisper their desires to robots. “Please build me this feature. Don’t make mistakes.” I’m having fewer meetings while working on more things. I went from typing all day to mostly yapping at my computer. Typing is now reserved for morning pages and this newsletter. It feels slower. More deliberate. Like thinking.

Almost everything about how I work, and how I define myself professionally, has transformed overnight. My career sits at the center of my identity, survival, and vocation, and it is undergoing a shift I can only describe as chaotic excitement and insane exhaustion.

I’m also having the most fun I’ve ever had in my career. That may be the better metric right now.

Chaos is an audit

Chaos exposes the beliefs, systems, and rituals that can no longer support the weight placed on them. It shows us what served the work and what survived because everyone agreed to keep doing it.

AI is producing a similar moment inside knowledge work. It is auditing expertise, titles, meetings, handoffs, approval chains, status updates, departments, and org charts. It reveals which activities create value and which merely perform the appearance of value.

Not every ritual was foolish. Some processes compensated for real technological constraints. Others gave people time to build trust, disagree, and arrive at a shared understanding. The work now is to separate coordination that helps people create from coordination that simulates progress.

Years ago, a manager repeated a phrase whenever things became chaotic: Execution over everything. Another mentor taught me to focus on outcomes. Together, they form a useful operating system: make the work real, then judge it by the change it creates.

From idea to outcome

For much of the software era, work was organized around distance. A designer lived in a canvas made of pixels, flows, and prototypes, then handed the work across a gap to an engineer who rebuilt it in a medium the designer could not speak.

Intent became a mockup. The mockup became a specification. The specification became a conversation. The conversation became code, and the code became something almost, but not exactly, like the original idea. Translation leaks, and the industry learned to live with the loss.

We built careers and management systems around that distance. Product managers translated between functions. Designers made representations of the thing. Engineers made the thing. Leaders reviewed abstractions because interacting with the real thing was expensive.

AI has blurred the gap. A designer can ship a pull request. I shipped three in my second week on the job. A writer can build a working prototype. A product manager can enter a meeting with something people can use rather than a document describing what might exist. The handoff loses authority when one curious person can move from idea to interface to code to feedback before the traditional team has finished scheduling the kickoff.

The titles are breaking down because behavior is outpacing our ability to name it. Titles are useful because human coordination is hard and labels save brain energy. They are also made up. An org chart is a story about how work happens. It remains useful until reality stops following the plot.

For most of my career, nonconformity was a liability. I did not go to a famous school, and my résumé wandered through freelance writing, content marketing, an education startup, community building, brand strategy, design, and management. The old system looked at that path and asked, “What do I do with you?” Now the same wandering path feels less like a defect and more like preparation.

The advantage belongs to people who can learn across disciplines, develop mastery in a few, and carry a good idea across the borders between them.

Execution is judgment made visible

When production becomes cheap, output becomes weak evidence.

Anyone can generate a strategy, an interface, an article, a research report, a prototype, or a pile of code. We are entering a world of infinite drafts produced at nearly zero cost, which moves the valuable questions up a level: Should this exist? Is it any good? What deserves more attention? What should we refuse to make?

I have watched non-designers use AI to make design-y things for the first time. They light up like a toddler hearing a recording of their own voice: an animated HTML deck, a button that wiggles, a little app that works. That delight is real, and so is the learning. But when something serious has to get made, the gap becomes obvious. The tool can produce options. It cannot give someone the vocabulary, pattern recognition, or judgment they have not yet earned.

This is why so many conversations about the future of work return to taste. Taste is the ability to choose. It knows which detail deserves three hours, which deserves three seconds, when the first answer is average, and when something is technically correct but spiritually dead.

Taste without execution remains an opinion. Execution is judgment made visible.

Execution means accepting responsibility for the path from intention to consequence: choose a worthwhile problem, define what good looks like, make something real, expose it to feedback, improve it, and stay with it until the desired change occurs. The outcome is the evidence. A strategy no one believes, a prototype no customer uses, and twenty projects in motion are possibilities, not outcomes.

What speed cannot replace

Some parts of work deserve to disappear. I have a Slack bot that reads my team’s channels and gives me a morning brief: who is working on what, which projects arrived, where people are blocked, and which decisions need attention.

I don’t have to ask everyone for updates, and they don’t have to stop working to perform an update for me. Good. No one joined a company hoping to spend Wednesday morning explaining what they did on Tuesday afternoon. Yet we call this work.

The brand strategy process is more complicated. At an agency, strategy could take four to six weeks because every decision downstream depended on it. I interviewed executives, team leaders, and customers, studied the company and its ambitions, then turned all of it into a direction the team could believe and build from. It also took me time to digest this information and write something cohesive.

Today I have a Claude Skill trained on years of projects, patterns, language, mistakes, and decisions. I can enter a company, examine its documents, talk with its leaders, and produce several legitimate strategic directions in a day. That would have sounded absurd five years ago. It is also incomplete.

But because I have years of experience and pattern recognition, I can tell whether the tool is producing good outputs. When I notice it's bad, I can trace the root cause in the code, learn what’s not working, and keep testing. Without the pattern recognition, you’re essentially just driving in circles and calling it an adventure.

The old process was inefficient, but inefficiency was not its only feature. Interviews made people feel heard. Workshops gave teams permission to disagree. A foolish comment could unlock the room. People developed language together and became advocates because part of them lived inside the outcome.

AI can compress production, but it cannot eliminate the human need for participation. There is a loneliness in arriving at the answer before everyone else has had the chance to ask the question. You can produce a brilliant strategy in one day and still fail because no one believes it. You moved faster than the group could develop trust.

Enrollment takes time. Sometimes the meeting is theatre. Sometimes it is where people decide to care. Experience means knowing the difference.

Execution of what?

It is easy to mock people who resist AI. It is more useful to understand what they are protecting. A degree told them they had expertise. A title gave them status. A specialized tool separated them from amateurs. A career gave them dignity, a sense of belonging, and a story about the future.

Then a machine performed part of that work in thirty seconds.

Of course there is fear. A livelihood becomes a private mythology made of education, sacrifice, promotions, surviving bad managers, and a language only people inside the profession understand. Someone can recognize that a new system is faster and still resist it because the old system helped them understand who they were.

A beginner may adapt more quickly because they have less of their identity invested in the old process. They lack experience and taste, but they also have less to mourn and fewer habits to defend. The opportunity inside this chaos is to build an identity around learning rather than a tool with the shelf life of a banana. The durable skill is the ability to become useful again.

AI also removes friction, and friction once forced us to choose. An idea that required six months, a meaningful budget, and ten people had to survive serious scrutiny. Now a passing thought can become an active project before lunch.

Desire becomes operational. You can speak an idea into existence, generate the research, create the interface, assign the agents, write the code, revise the copy, and open three more directions before the first idea has earned your commitment. My twelve subagents are not merely a funny anecdote. They are a mirror of my ambition.

Execution over everything does not mean execution of everything.

When nearly anything becomes possible, restraint becomes part of execution. The scarce ability is knowing what to ignore, which good idea must be abandoned so the great one has room, and what should remain slow because the slowness produces something valuable.

I know where permanent acceleration leads. My ambition has put me on the treadmill before, and I know the humiliating realization of discovering that my own finger was pressing the button. AI gives ambition a faster engine, but it does not give ambition a steering wheel. That remains our job.

Make the work real

Work in B2B tech long enough, and you will meet someone who believes their job is to “wrangle stakeholders” for three months so the group can achieve clarity: thirteen meetings, seven documents, endless threads, no decision.

Big tech companies produced and normalized this behavior. Now, with this new technology, it’s challenging that way of working.

Then someone makes the prototype in an afternoon.

The prototype does not settle every question. It changes the quality of the questions because people can respond to something real rather than defend their interpretation of an abstraction. Instead of arguing about whether an idea might work, create evidence. Give people something they can touch, test, reject, and improve.

The point is not to kill process. Process must earn its keep. Some processes make good work possible. Others defend the old ways of doing things.

Execution also does not mean becoming a lone hero. The best work of my life happened with small groups of motivated people who trusted one another, respected each other’s craft, and cared more about the outcome than the credit. One person saw the system. Another understood the customer. Someone could build, someone could write, and someone noticed the detail everyone else missed. You find those people and lock arms.

Titles will change. Tools will churn. Job descriptions will become outdated before the recruiter finishes posting them. What travels is character and portfolio: who you are when the work becomes difficult, and proof of what you made possible.

Become known as the person who makes good work happen without the drama. You do not need every answer. You need to move the group closer to one. Create clarity. Make the first version. Tell the truth. Ask the inconvenient question. Notice the customer everyone forgot. Remove the idea that no longer serves the outcome. Turn words into works.

My labels are still struggling to keep up with my behavior: writer, strategist, designer, manager, photographer, builder, whatever comes next. I’m tired in a way that is difficult to explain because the exhaustion sits beside genuine delight. Every week, something I once considered impossible becomes a Tuesday afternoon.

The opportunity is larger than replacing jobs. It is the chance to redesign them, remove the theatre, preserve the human collision, and give people more agency over their attention, energy, and time. That possibility deserves more than panic and predictions. It deserves proof.

At the end of some days, I close the laptop and have no words left. The next morning, I open my journal and write by hand. It is slower, which is why it works. The machines can help me make nearly anything. The blank page and pen ask the harder question: What is worth making?

---