Almost everything about how we evaluate business talent is optimized for specialists. Job postings are stacks of credential keywords. Career advice says pick a lane and own it. LinkedIn rewards a single sharp title. The whole machine is built to sort people by function, and it treats the generalist — the person whose answer to "what do you do?" takes a full paragraph — as someone who couldn't commit.
I'm that paragraph person. Twenty years running a business where I didn't just manage every function — operations, brand, e-commerce, IT, finance, HR, real estate — but worked inside each one. For most of my career, the market read that as a quirk at best. "Jack of all trades, master of none."
People forget the full saying: jack of all trades, master of none, but oftentimes better than master of one. And AI is about to make the second half of that sentence the most important line on the org chart.
AI commoditizes exactly what specialists sell
Be blunt about what's happening. The thing a specialist sells is skilled production: the deck, the model, the campaign, the code. That is precisely the layer AI is compressing fastest. I've watched it from my own desk — marketing work that used to need a team, financial modelling I used to need help with, software my vendors quoted at $80,000 — all of it now produced by one person with genuine fluency in the tools.
What AI does not produce is the decision about what's worth building, the standard for what good looks like, and the alarm that goes off when a confident answer is wrong. Those come from somewhere else. They come from having done the work.
What generalist judgment actually is
Judgment gets talked about like a personality trait. It isn't. It's pattern recognition earned across functions — and it's why operators who've worked inside the whole business carry something specialists structurally can't.
In a real business, nothing stays in its lane. A marketing decision is an inventory decision — advertise a product you can't keep in stock and you've paid money to anger your own customers. An inventory decision is a staffing decision is a cash flow decision is a lease decision. The specialist optimizes their lane, and often does it brilliantly. But when the lane itself is the problem, you need someone who can see the whole road. Twenty years of being accountable for every lane at once is twenty years of practising exactly that.
It's also twenty years of confident wrong answers — from vendors, consultants, spreadsheets, and now from AI models — and learning to smell wrongness before it costs you. That instinct is the single most valuable thing I bring to a tool that produces confident output at industrial speed.
The blunt part
A few things I believe that the career-advice machine gets wrong, stated plainly.
Credentials tell you someone learned a function. They tell you almost nothing about whether that person will own a problem — chase it across departmental lines, into the warehouse or dig into the lease document or scour the log files, until it's actually solved. I'll take a problem-owner over a credential-holder every time, and after twenty years of hiring, I can tell you the two overlap less than you'd hope.
"Focus" has been over-sold. Depth matters — but depth in one function, with no working model of the functions on either side of it, produces locally perfect decisions that are globally wrong. Some of the most expensive mistakes I've seen came from excellent specialists doing excellent work on the wrong problem.
And the AI-era version of the advice is backwards. The standard line is that AI makes deep specialization safer — go deeper, be the irreplaceable expert. For some genuinely rare expertise, maybe. But for most knowledge work, the moat isn't depth of production skill anymore, because production is what the machines do now. The moat is knowing what to build, judging whether it's right, and owning the outcome across the whole system. That's an operator's moat.
What to do with this
If you're building a team: stop screening for lanes and start screening for owned outcomes. Ask candidates what they've built end to end, what broke, and what they did when it broke. The person who's done the work across functions, multiplied by AI fluency, now does the output of a small department — I'm living proof, and I'm far from the only one.
If you're a specialist: broaden on purpose. Learn the function upstream and downstream of yours. Your production skill plus a working model of the whole business is the strongest position on the board.
And if you're an operator who's spent years feeling like the market didn't have a box for you — the market is about to need exactly what you have. The execution layer is compressing. Judgment is the scarce asset now. The people who've actually done the work are about to matter more, not less.