It’s a Compression Event!
And in Compression, the Differentiators Become Very Clear!
When McKinsey & Company announced it had effectively deployed 25,000 AI agents and saved 1.5 million hours of work, the internet did what the internet does best: blew a chip board or two and declared 750 consultants extinct.
Then Ernst & Young (EY) and PricewaterhouseCoopers (PwC) pushed back on the idea of counting AI as “headcount,” and suddenly the narrative flipped.
“They’re doing the same thing,” critics said, “just quietly.”
Both reactions miss the point.
No, 25,000 consultants were not marched out of glass towers by a fleet of polite chatbots carrying severance envelopes. But yes — something very real is happening.
And it’s bigger than layoffs.
The AI Takeover (That Isn’t a Takeover)
McKinsey didn’t say, “We replaced 750 humans with robots who now enjoy coffee breaks and annual reviews.” They said AI tools saved 1.5 million hours.
That’s a productivity story.
Consulting firms have been here before. Spreadsheets once terrified accountants. Cloud collaboration spooked project managers. Automation software made interns question their Excel wizardry.
Each wave eliminated manual effort. None eliminated the profession.
AI is simply the next — and louder — tool in that progression.
Calling AI “25,000 employees” is clever branding. It sounds dramatic. It trends nicely. But it reflects usage scale, not a robot uprising in tailored suits.
The Pyramid Problem
Consulting runs on a pyramid.
Lots of junior analysts. Fewer managers. A small, polished layer of partners at the top.
Traditionally, the base does the heavy lifting:
Research
Data cleaning
Model building
Slide formatting at 1:43 a.m.
AI has looked at that list and politely said, “I can do that in six minutes.”
That doesn’t eliminate consulting. It compresses the bottom of the pyramid.
And here’s where it gets interesting.
When the mechanical work shrinks, so does the training ground. Junior consultants historically earned judgment by grinding through repetition. If AI handles 60% of that repetition, firms must rethink how they manufacture experience.
You can’t fast-track wisdom with a plugin.
Tasks vs. Jobs (Or: Who Gets Blamed?)
AI is excellent at:
Drafting memos
Summarising research
Structuring analysis
Finding patterns
AI is terrible at:
Navigating executive egos
Making judgment calls when data conflicts
Handling client politics
Taking responsibility when a strategy fails
When a transformation collapses, nobody calls the algorithm into a board meeting.
They call the partner.
Consulting firms aren’t paid for slides. They’re paid for judgment, speed, and — most importantly — accountability.
And accountability is still stubbornly human.
Will the Billing Model Explode?
Some argue: if AI does 60% of the work, clients won’t pay 100% of the fee.
Possibly. Margins may tighten. Junior hiring may shrink.
But consulting was never purely about billable hours. Clients pay for:
Confidence
External validation
Political cover
Speed
Someone to stand next to them when things go wrong
The industry collapses only if clients stop believing external advisors add differentiated value.
Which brings us back to the real theme.
This Is Not an Extinction Event.
It’s a compression event.
Compression reveals structure.
When intelligence becomes cheaper, clarity becomes more valuable. When analysis becomes abundant, trust becomes scarce. When drafting becomes instant, judgment becomes premium.
The professionals who win will not be the ones asking, “What can AI never replace?”
They’ll ask, “What becomes more valuable now that AI handles the mechanical work?”
They’ll:
Integrate AI instead of resisting it
Move up the value chain
Develop judgment faster
Build relationship capital deliberately
Own outcomes under pressure
Because here’s the uncomfortable truth:
AI doesn’t destroy strong value.
It exposes shallow value.
And that’s why the debate between McKinsey, EY, and PwC isn’t really about headcount.
It’s about positioning.
Everyone is deploying AI. Everyone is restructuring around it. Everyone knows the pyramid is shifting.
The only real question is who adapts fastest.
This isn’t consultants vs. machines.
It’s professionals with judgment + AI versus professionals without either.
It’s not a takeover.
It’s a tightening.
And when things tighten, the difference between average and exceptional becomes painfully obvious.
As I call it, It’s a compression event. And in compression, the differentiators become very clear!
Comments welcomed, as always!
MC
