The scarcest skill in the economy just changed
It's not coding. It's not prompting. And you've been building it for 20 years.
I’m writing this from seat 3C on a Delta flight to Chicago.
An hour delayed.
I’m using the time to plan, outline, and write this issue, chatting with my AI writing partner between interruptions.
This newsletter used to take three to four hours of focused writing time. Today it’s happening in stolen moments sitting on the tarmac.
Also on my mind: Rory McIlroy winning the Masters last weekend by a single stroke. A week that looked like a heart monitor. Six-shot lead after two rounds. Tied after three. Down by three during the final round.
Then up by two going into 18, drove it way right, almost into the 10th fairway, bogeyed the hole, and still won by one. Reminds me of entrepreneurship. As long as you keep playing, you have a shot.
I can write this newsletter from an airplane seat because I know what a great newsletter looks and feels like.
Two decades of writing.
Thousands of newsletters sent.
I didn’t learn that last month.
I earned it over years of reps.
AI didn’t give me that knowledge.
But it made that knowledge explosive.
And everyone is getting something wrong right now.
The Replacement Myth
The dominant narrative: AI is coming for your job. It’s coming for the experts next. The robots will replace the humans.
I don’t buy it.
Not because AI isn’t powerful. It is. I use it every day. I’m using it right now.
But the narrative misses the point.
The bottleneck has flipped.
For decades, the constraint was execution. You knew what good looked like. You just couldn’t afford to build it. Couldn’t find the right people. Didn’t have the bandwidth. The ideas were there. The resources weren’t.
Now execution cost is trending toward zero.
But someone still has to define what “great” looks like. Someone has to write the spec. And that takes something AI doesn’t have. Years of reps. Pattern recognition earned in the field. Hard-won taste built one mistake at a time.
The real risk isn’t being replaced by AI.
It’s being replaced by a competitor who has expertise AND AI.
Define the What
I’ve been testing this in my own business. Let me show you what I mean.
Start with SEO. I’ve known what good SEO looks like for years. Never had the bandwidth to do it myself. Never had the budget to hire someone for something with a long payoff period.
Last month I built an SEO-optimized site in 48 hours (really about 3 hours over 2 days). We’re already seeing a 20% bump in traffic. I defined the what. AI built it.
We manage book projects for clients. Highly complex. A repeating production process, not individual projects. We’ve tried every project management tool. None fit.
Last year we spent five figures and three months building a custom Airtable app. It got us 60% of what we needed. I rebuilt it as a custom app in less than a week. Minus the $10K per year Airtable subscription. It’s at 80% now and improving every week. Updates are as easy as telling AI what to fix next. By voice.
Our LinkedIn ads used to eat three to four hours of my week. Now it’s 30 minutes. Another 30 to create new ad tests.
We went from testing monthly (because that’s all I had time for) to testing weekly. Cost per lead dropped from $846 to $532. Still trending down. I defined what good ads look like and what good performance means. Built that definition into the AI. I still make the calls. AI does the rest.
Then there’s this newsletter. I have an AI junior writer who interviews me for each issue. Organizes the outline from my ideas. Creates a first draft that I tweak and edit.
The finished product you’re reading could not exist at this level without my ideas, my thinking, and my definition of what a great newsletter looks and feels like. Built over two decades and thousands of newsletters sent.
The pattern across all four: I didn’t ask AI “what should I do?”
I told it “here’s what great looks like. Now build it.”
I do the highest-level thinking and decision-making. AI does the rest.
The Skill Nobody’s Talking About
You don’t need to be a programmer. You don’t need to understand machine learning.
The advantage is experiential.
If you’ve spent 10 or 20 or 30 years in your domain, you already own the asset. You know what great looks like. You’ve built taste that can’t be Googled.
That knowledge is now the scarcest resource in the economy. Not data. Not code. Not compute power.
The ability to define a great outcome.
And the separation is about to accelerate. Those who embrace this will compound. Those who cling to the old ways will fall behind. Not slowly. Fast.
Because unique IP begets more unique IP. Every time you define what “great” looks like and AI executes it, you learn something new. You refine. You get sharper. The muscle builds.
It’s not a gift. It’s a skill. And every rep makes the next one faster.
The Real Question
Forget whether AI is coming for your job.
Ask a better question: have you spent enough time doing the work to know what “great” looks like?
If you have, this is your moment.
Stop debating replacement. Start defining outcomes.
The world belongs to the people who can define the what.
See you next Saturday,
Steve “define the what” Gordon
P.S. I landed in Chicago and finished this between the gate and the Uber. Two years ago that sentence would have been impossible. What are you sitting on that AI could amplify, if you just defined the spec?
P.P.S. When you’re ready, our team is standing by to help you in three ways:
Become a Published Author — Million Dollar Author helps you extract your expertise into a book in 90 days. [Learn more →]
Want a Freedom Firm?: Read The Freedom Firm Manifesto >> The roadmap for a business and life you’ll love.
Watch: What Amazon’s 30,000 Layoffs Reveal About the Future
Where’s Steve:
April 19-21 - Exit Planning Summit, Nashville (at Gaylord Opryland)
May 7 - Small Business Expo, New York City (speaking)
May 7-9 - NAPFA Spring Conference, Minneapolis (speaking)
May 27 - Small Business Expo, Boston (speaking)
I’d love to meet you face to face. If you’re at any of these, come find me and say hello.