AI Junior Engineer: AI Version

Anthropic. (2026). Claude Sonnet 5 [Large language model]. https://claude.ai
Requested contextual analysis

So what are the differences between all three documents, in a nutshell

In a nutshell:

  1. Original — Your raw draft, unedited. All typos and grammar quirks intact (NetBUI, “electician,” “mathmatics,” “ompinpresent,” “aricle,” missing words like “now” instead of “know”). Full authentic voice, including the incomplete/run-on sentences.
  2. Grammar-corrected version — Same structure, same sentences, same voice — I only fixed spelling and grammar (NetBEUI, electrician, mathematics, omnipresent, article, “I now have”). Nothing rewritten, nothing reordered, nothing added or cut. Closest thing to a proofread copy of your original.
  3. Contextualized/rewritten version — Same ideas and same anecdotes (your dad’s switching story, the tech nostalgia list, the ache about junior engineers, the Ethernet point, “I added, I did not subtract”), but restructured for flow: smoother transitions, some run-ons broken up, a few phrases tightened or reworded for clarity, minor connective tissue added between paragraphs. This is the one where I made editorial judgment calls, not just corrections — which is also why it’s the one that needs the AI attribution note.

So: #1 is you, unfiltered. #2 is you, spell-checked. #3 is you, edited — meaning is preserved but the sentences aren’t all originally yours.

AI is the best junior engineer I’ve ever had.

I’ve been in IT for over 30 years now, and everyone is all aflutter about AI replacing their jobs. There’s truth to that. AI has already replaced plenty of the tasks I used to do by hand.

But here’s the thing:

My dad joined the Navy and became an avionics electrician. He parlayed that into a job at the phone company, from which he retired 30 years later. He started out with physical electric switching equipment — actual tiny pieces of metal making contact with each other to complete a phone call. By the time he retired, he still knew how to fix those old physical switches, but through training in the back half of his career, he’d mostly moved to issuing commands into computers that did all that switching digitally, out of physical view and out of his hands. I remember that shift. I remember the training that came with it — weeks-long stretches where he was gone, and my mom and I would sleep in the same room because it made us both feel better. But he survived it, and he understood it. And telecommunications still basically work the same way today: you dial some number or ID, and you talk to the person, or you don’t. Video’s in the picture now, and the how has changed completely — but the what hasn’t. Same function. Different modes and tools.

Now is no different.

My own 30 years: LOAD “*”,8,1 — hand-coding HTML in Notepad — configuring NT 3.1 — discovering Linux and making it work — coax cable — Web 1.0, 2.0, and everything after — social media — and now AI. Oh, and IPX/SPX and NetBEUI, anyone? All of that is a memory now. But the incredibly enabling technology I got to use was made possible by good people who came before me and figured out how to make it all work in the first place. Such is progress.

So — AI is the best junior engineer I’ve ever had. I’ve had my own fits with it. Ask me sometime how I burned $250 in one hour because I didn’t know what I was doing — that’s an article for another day. But I learned. And now AI is so enabling for me. It’s a force multiplier. Whatever I’ve ever wanted to build — server-side, code-side, idea-side — I now have an instant, knowledgeable engineer at my fingertips. So far, it’s excellent — but I have to guide it, mentor it. Left alone, it goes off the rails fast. It’s a junior engineer. The raw talent is real, but it still needs to be fostered and guided.

So where does that leave the actual human junior engineers of today? What about them?

This is the part that aches for me. I worry they won’t learn the foundations I had to learn, because AI will just do it for them.

And then it hits me: that’s exactly the point.

I understand a lot of IT, down past the network layer into the physical layer — but then I stop. Somebody before me figured out how to make Ethernet happen in the first place. (I know who, roughly — you know what I mean.) I never had to reinvent that wheel. I only ever expanded on it. And I thrived because of it.

Human knowledge evolves. I’ll add a small caveat there — we’ve all seen the Terminator movies — but my mind still settles into a kind of peace about it. This will be their world. Today’s junior engineers will do things I can’t even fully comprehend. I have an inkling: DNA medicine, materials science, pure mathematics (someone give us Star Trek warp drive already). That will be their world, built the same way mine was built for me — by standing on a foundation somebody else laid down, and finding it as natural as breathing.

Today’s junior engineer will be my guide for tomorrow. I say that with real deference.

As long as we keep knowledge flowing forward, and keep it free, we’ll be okay. (Education costs and access — that’s an article for another day too.)

This is an inordinately optimistic place for me to land. But it’s been sitting in my head constantly as I wrestle with the truth of it, and honestly, a touch of the harder feelings that come with it. I’m old. Not old-old — but old enough to know it. I’m not out of it, though. I’ve still got fight. I’ve still got moxie. I’ve still got knowledge. And I don’t think it’s too much to ask of myself to focus on what I know, adapt, and guide myself toward whatever the next milestones turn out to be — content in knowing I was part of the whole process.

I added. I did not subtract.

Proudly powered by WordPress