The New Math of Work Where Do You Want to Sit?

A small fraction of your people already do most of the work that matters. AI is about to shrink that fraction further - and make the people inside it worth far more.

June 12, 2026 8 min read
The square root 80/20 Resistance to AI Putting it together How to be in the group Where this leaves you

A small fraction of your people already do most of the work that matters. AI is about to shrink that fraction further, and make the people inside it worth far more. Taken to its edge, the idea is almost uncomfortable: a minority of your headcount, the high performers who use AI well, could carry close to the whole company.

That sounds like an overstatement. Walk through the numbers and it stops sounding like one. Three separate observations, each well documented, point the same way. Where do you want to be standing when they land?

A long dark conference table lined with identical empty chairs receding into shadow, with a single chair pulled into a warm pool of light - the small fraction that carries the work, and the question of where you want to sit

The first number: the square root

In the 1960s, a physicist and historian of science named Derek de Solla Price noticed something while studying academic publishing. Roughly half the papers in any field came from the square root of the contributors. Fifty authors? About seven of them produced half the work. Ten thousand researchers? Around a hundred carried half the load.

This became known as Price's Law. People have since stretched it well beyond academia and applied it to organizations in general. In a company of 100, maybe 10 people do half the work. In a company of 10,000, perhaps 100 do.

100 → 10

In a company of 100, about 10 people do half the work.

10,000 → 100

In a company of 10,000, around 100 carry half the load.

Price's Law started as an observation about scientific authorship, and the exact square root won't hold in every setting. The ratio matters less than what it captures: output is lopsided, and it gets more lopsided as organizations grow.

The second number: 80/20

You already know this one. The Pareto principle says 80% of results come from 20% of inputs. It shows up in sales pipelines, bug reports, customer revenue, and yes, employee output. The split is rarely exactly 80/20, but the shape repeats everywhere you look. A minority of effort drives a majority of outcomes.

The shape that repeats everywhere
Inputs · the effort, the people20%
Outputs · the value, the results80%

A small slice of input produces most of the output. Rarely exactly 80/20 - but the shape holds across sales, revenue, bugs, and employee output alike.

Price and Pareto aren't the same measurement, and they don't stack on top of each other. They're two readings of one phenomenon, and they agree. Productivity is concentrated. Somewhere between "the square root does half" and "20% do 80%" sits the real distribution in your organization.

The third number: resistance to AI

Now add the variable that's reshaping all of this. AI is landing in workplaces fast, but it's landing unevenly, and a meaningful share of people are pushing back. The numbers depend on how you ask, but the picture holds up.

0
of CEOs say most employees are resistant or even openly hostile to AIKyndryl, 2025
0
of employees admit to actively undermining their company's AI strategyWriter, 2025
0
of US workers say they never use AI in their role at allGallup, late 2025

So "roughly a third actively resist" is a defensible read of the landscape, sitting somewhere between passive avoidance and active sabotage. Call it 30% as a round number, but hold it loosely. The more accurate version is "a substantial minority that isn't going anywhere fast."

Putting it together

Combine the three. A small group already produces most of the value. A sizable minority resists the biggest productivity multiplier to come along in a generation. The people who both do the work and use AI well end up with an outsized share of the impact. Follow that to its end and a fraction of your headcount, the high-output people who lean into AI, could run something close to the whole operation.

That last step is a hypothesis, built on top of patterns that are not. Price and Pareto have held up for decades. AI resistance is measured and current. What's new is the intersection: the high performers who adopt AI well are pulling away from everyone else, including from the high performers who don't.

Output → high
Coasting

High output, AI holdout

Still a strong performer - but getting passed by peers who do the same work with AI. The lead erodes quietly.

Pulling away

High output, AI-fluent

The rockstars. Already carrying an outsized share, now compounding it with tools the resisters won't touch.

Left behind

Low output, AI holdout

Neither the work nor the multiplier. This is the group the new math sorts out the fastest.

Rising

Lower output, AI-fluent

Punching above their old weight. The multiplier lets them close ground on the coasting holdouts.

holdout ← AI adoption → fluent

These are the people becoming the organizational rockstars, the ones already carrying an outsized share of the work and now compounding it with tools the resisters won't touch. The compounding is real and large.

0
gains we are seeing at Strongly - a week of work done in a dayengineering, analysis, sales outreach
0
wage premium for workers with AI skills - more than double the year beforePwC, 2025

The gains are uneven, and that is the whole point. Some people capture far more of the upside than others, and that gap is the rockstar effect. AI doesn't lift everyone equally. It widens the distance between the people who wield it well and the people who don't.

The market is pricing this in now. So the question stands. Where do you want to sit?

How to be in the group that's excelling

The good news is that this isn't a fixed caste system. Both the high-output group and the AI-fluent group are earned, not assigned. Here's how to put yourself in them.

Move from output to outcomes

The people who drive disproportionate value rarely just do more. They do what matters. Find the 20% of your work that creates 80% of the value, and protect time for it. Busywork is where careers plateau.

Adopt AI deliberately, not superficially

Most so-called AI users only search, summarize, draft an email. The people pulling ahead rebuild whole workflows around it. Pick one real workflow and rebuild it around AI. Then pick another.

Pair AI fluency with judgment

AI raises the floor on producing something, not on producing something good. The differentiator knows which output to trust, which to throw away, and what to ask next - and makes AI work for the whole team.

Be a critical adopter, not a holdout

Skepticism is an asset: it catches errors and flags risks. Flat refusal to engage is what sorts people into the group left behind. Push the tools hard, and trust nothing they produce without checking it.

Build the habit before it's mandatory

The wage premium and the job security go to the people who started early, while it was still optional. The window where AI fluency sets you apart, rather than being the baseline, is closing.

If you lead a team

Know who your square root is and do not let them walk. Clear the friction that keeps your best people from adopting AI. And watch the resisters: the goal is to convert them, not to carry them.

For leaders

The same math is a to-do list. Every week your best people wait to adopt AI, the gap they could be opening for you stays closed - and the gap a competitor's best people are opening keeps widening.

Where this leaves you

The exact ratios will vary by company. The pattern underneath them does not. Value concentrates in a few people, AI is stretching the distance between those people and everyone else, and pay is already following.

The people opening the gap are doing it now, while it still counts as being early. The only real question is whether you are one of them.

References

Put your best people on the right side of the gap.

Strongly's forward deployed engineers embed with your team, rebuild the high-value workflows around AI, and leave your people fluent enough to keep compounding the gains. The early movers are pulling away now.

Explore the Platform