🧠 Should I double down on being a generalist now that there’s AI?

I recently received this question on socials:

“Given what’s possible with AI, do you think it’s important for data engineers to branch into software engineering too?”

Honestly, if you’d asked me this two years ago, I would’ve said no. A key thing in my career has been to niche down (data engineer → devrel in data engineering). But I’ve been building way more stuff outside of data engineering these past months.

So what’s the model can’t replicate? I think deep niche understanding about the business and domain expertise still matters—but the execution layer is getting commoditized. The question isn’t “should I learn React?” anymore. It’s “do I understand the problem well enough to direct an AI to build the right thing?”

🛠️ Cursor’s plan mode

I don’t know how I missed it, but back in October 2025, Cursor released its plan mode.

It enables the model to research your codebase to find relevant files, review docs, and ask clarifying questions before writing any code.

This is basically what I used to do manually (and it still works if you aren’t using Cursor’s plan mode). Whenever I’d start a new project where I wasn’t sure about the architecture, UX, or technical implementation, I’d use a prompt like this while pinning relevant docs I’d previously indexed within Cursor settings:

Don't start any coding implementation yet. I'm thinking about X—ask clarifying questions and use @docs to verify what's possible.

Now plan mode does roughly that, and the UX within Cursor is nice since I can easily edit the plan and give the green light to build.

📚 What I read/watched

AI’s trillion-dollar opportunity: Context graphs: Foundation Capital (VC firm from the Bay Area) discusses how the advantage in industry at the macro level is shifting from “systems of record” to where decisions are actually being made. Agents won’t replace systems of record, but the key missing piece is decision traces—recorded exceptions, approvals, and precedents—so agents can show why decisions were made.
Can a VFX Artist Beat AI?: Zach King, the “magician of the internet,” made a fun video that reminded me you can still be really creative with AI rather than letting it beat you.
The Thinking Game: A nice documentary about DeepMind (acquired by Google) and the breakthrough that would eventually win a Nobel Prize.
The Big Constraint Flip: I know I talked about this in my last issue, but here’s another signal that resonates with me: we just passed a flipping point.

OK, I don’t know how I missed that UFO, but Raye’s live performance of “Where Is My Husband” is absolutely INSANE. EVERY live version. Such musicians, such a performance.