I was watching the NVIDIA conference last week and Jensen Huang said something that made me pause the video and just sit there for a minute.
75,000 employees. 7.5 million AI agents.
I did the math three times because I thought I misheard him.
That's 100 AI agents for every single person on his team.
Not in some distant future. By 2036. Ten years from now.
I currently have maybe ten AI agents doing real work for me. Huang is planning 100 per person. I'm either way behind or he's way too optimistic. Probably both.
He wasn't being theoretical about it either. He said these agents will be "working around the clock so our people don't have to keep up with them."
Which is a polite way of saying humans can't compete with machines on speed anymore, so we better get really good at directing them instead.
And while I was still processing that, I found out McKinsey already has this running.
25,000 AI agents. 40,000 employees.
They're not waiting for 2036. They built it last year.
So I started looking into what these companies are actually doing with all these agents, and that's when I found the Cognizant report.
They analyzed 18,000 tasks across 1,000 different jobs. The kind of deep research that takes months and costs a fortune.
Their conclusion: 93% of jobs will be impacted by AI.
Not replaced. Impacted. Which is consultant-speak for "your job is about to change whether you like it or not."
They put a dollar amount on it too. $4.5 trillion in labor shifting from humans to machines.
That's trillion with a T. For context, that's roughly the entire GDP of Japan. Just casually moving from people to computers.
And then they admitted something that actually surprised me. They said this is all happening six years ahead of schedule. What they projected for 2032 is already here.
Which means the timeline everyone's been working with is wrong.
So now I'm sitting here running five brands, and I'm thinking about what 100 AI agents per person actually means.
Right now I've got AI agents handling research, writing drafts, editing copy, pulling analytics. Maybe 10 agents total doing real work across everything.
Huang is talking about 100 per person.
That would mean I could theoretically run 50 brands at the same scale I'm running five right now
The question I keep coming back to is this: what happens when your competitors figure this out before you do?
Because they will… Some of them already have.
You've got two paths here. You can be the person who gets replaced by 100 agents, or you can be the person who commands them.
Most people are still thinking about AI as a helper, a chatbot, a thing you use when you're stuck on something.
That's not what this is turning into.
AI agents are becoming your workforce. They don't sleep. They don't need breaks. They don't have bad days. They just need direction.
And the gap between people who understand this and people who don't is going to get ugly fast.
Remember when "I'm not good with computers" was a funny personality trait? Yeah, that's not going to age well.
Huang said every company needs an "agentic system strategy." He compared it to HTML and Linux. The foundational stuff you can't avoid if you want to stay relevant.
If you're treating AI as optional right now, you're making the same mistake Blockbuster made when they passed on buying Netflix for $50 million.
Blockbuster is now a Twitter meme and a nostalgic memory. Netflix is worth $300 billion.
The companies that win over the next five years will be the ones building their AI workforce now.
Training it
Deploying it
Letting it handle the repetitive work while humans focus on decisions only humans can make
The companies that lose will be the ones that kept waiting, kept hoping this would slow down, kept thinking they could outwork a machine.
You can't.
So stop trying to compete on effort. Start competing on leverage.
Build your team of agents. Train them. Deploy them.
Look, I get it. This sounds overwhelming. 100 AI agents… Most people don't even know where to start with one.
So let me make it simple.
A few weeks ago, I built my AI assistant in 10 minutes. No code, and no technical setup. I recorded the whole thing and put it on YouTube.
And if you want to go deeper, I also built a full dashboard that runs all five of my brands.
Real-time analytics.
Content pipeline.
AI agents doing the grunt work.
I wrote the complete breakdown in a recent newsletter.
This is the new org chart. 100 agents for every human.
Are you building yours?
—Navin
