2 important things happened in the last 48 hours.

People are reading them separately — they shouldn't be.

Yesterday, Anthropic published the most detailed picture we've ever had of what AI is actually doing to jobs. One million real conversations. Real task data. Real success rates.

This morning, the February jobs report quietly confirmed what a lot of us are already feeling in our bones. The labor market we thought we had last year? It wasn't quite the one we actually had. Job growth was overstated by nearly a million positions.

I thought about both of these for a while this morning.

And what I keep coming back to isn't the scary part, it's something most people aren't thinking about.

Here's what the Anthropic data actually found.

The sophistication of your prompts almost perfectly predicts the usefulness of what AI gives back. The correlation was r > 0.92 — essentially a straight line.

The report frames this as a human capital finding. People who think more clearly, ask better questions, and know how to evaluate what comes back — they extract dramatically more value from AI than people who don't.

AI fluency matters. But it connects to something.

How clearly you understand a problem. How well you can articulate what you need. How quickly you can tell when the output is wrong.

That's the real variable. And it has a name we've been questioning for over twenty years.

Emotional intelligence.

Let's be clear it was never just a "nice to have." Or just another leadership buzzword (though we could fill a book with those).

Today, it's the skill that determines how much value you can actually extract from every AI tool you'll ever use — and the thing AI still genuinely, structurally cannot replicate.

Here's what I mean.

AI can write the report. It cannot read the room when you're presenting it.

It can draft the message. It cannot sense that the relationship is more fragile than it looks on paper.

It can synthesize the data. It cannot tell you that the real issue in the room hasn't been named yet, and that naming it is the only thing that will actually make a difference.

I've spent years coaching people through some of the most complicated professional transitions imaginable. And what I keep seeing, in the people who succeed, who build something that lasts, who stay relevant when everything around them is shifting, isn't the latest buzzword in that book.

It's the stuff that's inherently human.

There's a reason history repeats itself. Our brains haven't changed much over all those millennia — only the circumstances and surroundings have.

AI is just the latest variable.

The capabilities that helped people navigate every major disruption before this one are the same ones that matter now. We just finally have the data to prove it.

Those capabilities don't just survive an AI economy.

They compound in one.

Every year AI gets better at the analytical layer, the human layer underneath becomes more valuable. Not less.

So here's your playbook. Four things worth building right now.

  1. Get honest about how you react under pressure. Not to judge yourself — know yourself well enough to choose your response instead of just having one. That one second between trigger and reaction? That's where most outcomes are actually decided.
  2. Practice reading what isn't being said. AI processes what's in front of it. It can't sense the shift in emotion when a conversation changes, or hear what someone is leaving out. You can. Most people don't work on this skill and it's never mattered more.
  3. Build real trust. People are more anxious than ever (with reason). The people who can have an honest conversation without triggering defensiveness, who give feedback that's useful — those are the people others want around them, recommend, and protect. This isn't a personality type. It's a skill.
  4. Know when to put the AI aside. This one doesn't get talked about enough. The most strategic skill right now isn't knowing how to use these tools. It's knowing when the moment requires something they simply can't give. Presence. A real decision. A human being willing to own it.

The jobs data this morning was (very) uncomfortable.

Not because it was catastrophic. Because it wasn't.

It's because it's part of a quiet, continuing confirmation that the math is changing faster than most people are letting themselves acknowledge.

The professionals who manage this well won't be the ones who accumulated the most AI certifications.

They'll be the ones who invested in the human capabilities whose value builds over time. The ones that silently sit underneath every tool, every role, every successful change.

The ones that were always going to matter, we just didn't have the data to prove it until now.