The Skills That Survive AI

I’ve had more conversations about career anxiety in the last month than in the last three years combined.

Inside my team. Outside my org. Friends in data, engineering, product. The question is always some version of the same thing: “Is what I do still going to matter?”

It’s a fair question. When I started my career, I thought knowing SQL and Python made me valuable. And it did - for a while. Then everyone learned SQL and Python. The tool stopped being the differentiator. The thinking did.

We’re in that same moment again - just faster and at a much larger scale.

Here’s what I’m actually seeing across my teams right now:

• Our data scientists are shifting from writing one-off queries to building AI agents that do the analysis autonomously. The job is no longer “answer the question.” It’s “build the system that answers questions and gets better over time.”
• Our engineers are using agentic coding tools to ship faster. System design, architecture, judgment calls - still entirely human. But the implementation layer is compressing fast.

The work is changing. That’s not a prediction - it’s already happening.

So what separates the people who are thriving from the people who are stuck?

It’s not who adopted AI tools the fastest. It’s not who’s the most enthusiastic about the technology.

It’s the people who do both things at once: they adopt fast and they know the limits.

The most dangerous person on a team right now isn’t the one who refuses to use AI. It’s the one who’s confidently wrong with it - who ships AI-generated outputs without understanding where the model breaks, where the context is incomplete, or where the judgment still has to be human.

The people I’d bet on are the ones who: -> Treat AI as a system to build on, not a tool to use -> Know what the technology can’t do as clearly as what it can -> Build leverage - systems, context, frameworks - not just deliverables -> Understand that their value was never the output. It was the judgment behind it.

This is the same shift I see in data science and GTM - the future belongs to people who build systems that reason and act, not people who run the same playbook faster.

The anxiety is real. But if you’re feeling it, it means you’re paying attention. The people who should worry are the ones who aren’t anxious at all.