Where have all the Engineers gone?

This is part one of a short mini series. Part 2 is Rise of the Agent Director. Part 3 is You Can’t Prompt Your Way to being a Senior Engineer.

Where is my John Wayne?
Where is my prairie song?
Where is my happy ending?
Where have all the cowboys gone?
Where is my Marlboro Man?
Where is his shiny gun?
Where is my lonely ranger?
Where have all the cowboys gone?
Where have all the cowboys gone?
Where have all the cowboys gone?

Lyrics by Paula Cole. If you don’t know the song, go ahead, watch here and keep reading ;*

AI is taking over the world! Skynet is here, and we must get ready to serve our overlords…

Ok, maybe not that extreme, yet, but AI is slowly taking over. Engineers across the globe are adopting AI into their day-to-day, some much faster than others. I agree, we must adopt it, this is the new era of technology and those not keeping up will get left behind.

In my own experience, and what I’m hearing from others, the code from AI is good, but rarely acceptable as is. The inline completion and peer programming that CoPilot and Cursor offer have been reliably good, saving us a second or two per line of code, over a story that can amount to hours saved. Autonomous coding agents are where the code starts to vary and become inconsistent with what we like to see. Our experience is required to ensure that the agent’s changes are the right changes.

---

Subscribe to my blog and get posts like this in your inbox. Share your email below, or follow me on Threads, LinkedIn, or BlueSky.

---

Where is my John Wayne?

As a result, we frequently say our coding agents are “junior engineers”, or the more politically correct label: “early career engineers”. We have to give them (juniors and agents alike) clear requirements, look over their shoulders, tell them what to fix, then review all their code; rinse and repeat. In this capacity, it takes an experienced senior engineer to maximize the value of our coding agents, because they know what to look for.

Here’s my dilemma: If it takes a senior engineer to recognize poor code and guide someone to create acceptable code, and me, as a junior engineer, am using AI just as much as the seniors are, will I ever mature into a senior engineer? Time doesn’t dictate engineering seniority; experience does.

We’re going to explore this dilemma over my next few posts.

Where is my prairie song?

Our current junior engineers will inevitably become the next generation of senior engineers. I don’t plan on working forever. Junior engineers are coming up in a fundamentally different environment than seniors have; one where AI plays a central role in coding, debugging, and even architectural decisions.

And yet, we still need engineers who can write and understand good, clean, maintainable code. That foundational craft doesn’t disappear.

This creates a tension: senior roles are evolving from hands-on code to decision-making, orchestration, and system-level thinking. But those decisions are only sound when grounded in hands-on experience.

Early-career engineers need intentional friction without AI.

Junior engineers will miss the foundational learning experiences when they rely too heavily on AI tools.

• Troubleshooting complex issues independently, getting frustrated, failing, and learning.
• Working through the nuances of specific languages and their syntax.
• Thinking critically about performance and security, truly creating “good” code.
• Reading technical documentation, yes, we used to do that.
• Learning through trial and error, and trial and error, and trial and error, and…

This is what helps us gain experience and intuition: failure. (I’ve learned the most fighting a bug for hours on end, and my lack of hair proves it.) This is why good senior engineers become religious about certain patterns and methodologies, because they have earned it, they have tried various approaches and found the best ones that work for them. I love hearing my teammates struggling with issues because I know when they win and get through it, they are better for it, and our team is better for it.

These hands-on experiences foster deep understanding and skill development, which are essential for becoming a senior engineer.

With these incredible AI tools, debugging sessions are a matter of telling AI what the problem is and letting it figure it out. We’ve lost the edge. Failure is so much harder now.

Where is my happy ending?

I don’t know.

Here’s what I’m still struggling with:

• How will junior engineers develop critical thinking and problem-solving skills without direct, hands-on experience?
  • Can we give them these growth opportunities? Should we refuse them access to highly efficient AI tools for the sake of growth and preparing for the future?
  • Can we effectively backfill retiring senior talent if we don’t proactively support junior development?
• Is it time to redefine what “senior engineer” means in the age of AI?
  • If AI handles most coding and debugging, is the senior role more about decision-making, systems thinking, and aligning tech with business outcomes?
  • Similar to airplane pilots who rely on autopilot but must respond skillfully in emergencies, do senior engineers become the calm in the chaos (land that plan upside down)?
• Are we training engineers to think deeply, or just to generate and evaluate AI outputs?
  • Can product teams wait for engineers to grow into the critical experienced senior engineers we need? Or do we always need to ship it faster?
• Will traditional engineering roles still exist in a few years?
  • Is this even a concern? Will AI get “that good” at coding, we don’t need to review it, and senior devs are all about the system and architecture?

I don’t know. What do you think? I know there’s a tidal wave of a change coming, and I want to be ahead of it.

In my next post we’ll explore how engineering is changing with the advent of agents and what that might look like very soon. Later, we’ll explore what we might be able to do to address these concerns.

Yippy-yi, yippy-yay