Morning rant, AI is almost free ...
So, it only makes sense, it's simple math ...
Business value = Output gained β AI cost β operational risk
And when output concentrates around a few AI-enabled employees:
Future org design = fewer people higher leverage better tooling
Every prompt has a cost. Input tokens cost money.
Output tokens cost money.
And for anyone unfamiliar with tokens, one token does not equal one word. A single word can be multiple tokens.
But here is the business math most people are avoiding.
Imagine a company has 10 employees in one business unit.
Historically, that group produced a certain level of digital output across accounting, marketing, support, operations, or any number of corporate functions.
The company gives all 10 employees AI tools.
Productivity increases 5x.
But when leadership looks closer, they realize almost all of that extra productivity is coming from only 2 of the 10 employees.
Those two learned how to use AI well.
The other eight mostly kept working the old way.
At that point, the business question becomes uncomfortable but obvious:
Why keep funding 10 seats of labor when 2 AI-enabled employees are producing most of the incremental value?
Even if the company spends $50,000 per year on AI usage for each of those two employees, that is still tiny compared to carrying eight employees at $145,000 each, plus benefits, overhead, management time, office costs, and the fact that humans do not operate 24/7.
This is the part of AI disruption people do not like to say out loud.
AI will not replace everyone equally.
It will amplify the people who know how to use it, and expose the productivity gap around those who do not.