Andrew Ng just drew a line in the sand for engineers.

In an interview a couple of days ago, he laid out a hierarchy of engineering talent in the AI era:

1️⃣ Top tier: Experienced engineers (10–20+ years) who’ve gone all-in on AI and use it as a force multiplier. They move faster than anything we've seen.

2️⃣ Next tier: New grads who really know AI because they grew up with it. They are using tools, agents, and models as default, not exception.

3️⃣ In trouble: Experienced developers who had a comfortable job and are still “coding like it’s 2022” as if AI never happened. Ng says he just doesn’t hire people like this anymore. 🫥

4️⃣ At the bottom: computer science grads who never learned AI at all. They’re already behind the market.

This isn’t just Ng. It’s consistent with what we’re seeing across the industry:

Brian Armstrong has fired people who refused to use AI tools and couldn’t justify why.
Google leadership is telling engineers to adopt AI-assisted coding to stay competitive.

In my own work with agentic coding, I’m seeing a similar split: Some senior engineers are becoming AI-native builders – orchestrating fleets of agents, layering quality gates, and shipping at 10–20× their previous throughput.

Others are basically treating AI as a gimmick or a toy, then wondering why they feel threatened by it.

If you’re an engineer, three honest questions to sit with:

Are you still coding like it’s 2022?
Are you actively learning how to design workflows around AI agents, not just sprinkling in Copilot/ChatGPT here and there?
If someone observed your day for a week, would they say, “This person is in tier 1 or 2 of Ng’s hierarchy”… or tier 3/4?

And if you’re a leader:

Are you hiring, promoting, and rewarding AI-native ways of working?
Or are you still writing JDs and running interviews as if this shift hasn’t happened?