AI is taking over the world! Skynet is here, and we must get ready to serve our overlords… Ok, maybe things aren’t 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. not keeping up will get left behind What does this all mean for engineers? Currently, we need real experience, senior engineers, to review and iterate with AI to generate good code. What about those who are just starting off? What happens when the old folks retire? Where have all the Engineers gone? 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? 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 ;* Lyrics by Paula Cole. If you don’t know the song, go ahead, watch here and keep reading ;* watch here and keep reading 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. Our experience are the right changes 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. Our current junior engineers will inevitably become the next generation of senior engineers. 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… 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. failure 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 so that we don’t need to review it, and senior devs are all about the system and architecture? 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? 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? 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 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? 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)? 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)? If AI handles most coding and debugging, is the senior role more about decision-making, systems thinking, and aligning tech with business outcomes? 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)? 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? 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? 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? 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 so that we don’t need to review it, and senior devs are all about the system and architecture? Is this even a concern? Will AI get “that good” at coding so that we don’t need to review it, and senior devs are all about the system and architecture? Is this even a concern? Will AI get “that good” at coding so that 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 change coming, and I want to be ahead of it. Yippy-yi, yippy-yay Yippy-yi, yippy-yay Rise of the Agent Director Just as the rise of automation and CI/CD pipelines reshaped operations teams by forcing traditional sysadmins to upskill into DevOps engineers or risk obsolescence, we’re seeing a similar inflection point with AI in software engineering. The expectations of engineers are changing fast. Those who adapt and learn to manage agents will thrive. Those who don’t may find themselves deprecated, irrelevant, and obsolete. This is not because they weren’t smart, but because they didn’t evolve. “The quicker you let go of old cheese, the sooner you find new cheese.” Spencer Johnson, “Who moved my cheese?” “The quicker you let go of old cheese, the sooner you find new cheese.” Spencer Johnson, “Who moved my cheese?” We’re already seeing the shift. Engineers are building agents, experimenting with them in IDEs, integrating them into workflows, and discovering just how powerful and helpful they can be. Over the last month, I’ve created a video agent, a security vulnerability remediation agent (really rolls off the tongue), a ground truth analyzer, a legacy code remover for our 220ish repos, and more. In my free time, I busted out a new Star Wars site, a Star Wars meme app, and have 2 more (not all Star Wars) in progress. In just hours, I’ve been able to accomplish monumental tasks (and have a lot of lessons learned and battle wounds); it’s incredible. Engineers are falling in love with agents. I’m not falling in love, I’m already in love. Star Wars site Star Wars meme app This is a natural evolution, just like version control, open source, and DevOps have changed the shape of how we work; agent orchestration is the next evolutionary step, and will move faster than we’ve seen before. Unlike version control, open source, DevOps, etc., which have benefited the way we make software, AI changes the way we interact with how we make software. We don’t need to be super technical to make an app. Business is going to move tremendously faster because they don’t care about code; they just want their app. We, engineers, need to evolve, too. Evolve into an Agent Director Soon, we won’t just be coding with agents, we’ll be running teams of agents. And we’ll need engineers who can direct, supervise, and course-correct them when things go sideways, because they will. This is where experienced senior engineers come in, not as the best coders, but as the best judges of code, architecture, and trade-offs; decision makers for the agents. with of judges of code, architecture, and trade-offs AI can write it. You decide whether it’s right. AI can write it. You decide whether it’s right. This is the future, get ahead of the curve, now. Software engineers should begin to transition into a new role: Agent Director. I’m sure someone else will come up with a catchier buzzword. I truly believe this role is the future of engineering, and frankly, it is here today. We’re still early in the hype cycle, but it will catch on. It’s here, and it’s glorious. What is an Agent Director? Why, someone who directs agents, of course! :D An Agent director will be focused on: Creating, maintaining, and evolving agent configurations to optimize outcomes Monitoring agent performance and debugging the full pipeline, from prompt to production Training and fine-tuning agents to meet engineering and product standards Managing and observing workflows driven by AI Ensuring agents adhere to architectural constraints, compliance requirements, and security policies through well-defined documentation Validating high-level system design decisions And whatever else an agent can’t do, yet Creating, maintaining, and evolving agent configurations to optimize outcomes Monitoring agent performance and debugging the full pipeline, from prompt to production Training and fine-tuning agents to meet engineering and product standards Managing and observing workflows driven by AI Ensuring agents adhere to architectural constraints, compliance requirements, and security policies through well-defined documentation Validating high-level system design decisions And whatever else an agent can’t do, yet Huh, each one of those could be a blog post. Huh, each one of those could be a blog post. Some of this will be new: it’s clear there are new skills to be learned. We should be finding opportunities to learn this today. If you can’t do this at work, start a side project, or 3, and let the agents do the work. Some of this isn’t new and probably looks similar to what team leads are already doing today. It’s not all flowers and agents The agent director role comes with real risks. Plenty of folks have raised these, and I’ve mentioned them as well above. We’re a little early right now to truly appreciate the impacts, but it’s a matter of time. We have to be thinking about this now. Erosion of foundational skills can happen when we rely too heavily on AI, making us dumb. It may weaken our judgment and code quality instincts. There are some early studies out there in agreement. I would argue: if this is done right, this frees us to do more creative things, stretching new parts of our brains. But I’m an optimist. Loss of system and architectural consistency. Today, agents run great locally, can optimize code within their immediate view, but frequently miss systemic design principles and the “bigger picture”. As context windows get larger and models get smarter, will this become a non-issue? If we implicitly trust AI and assume agents are “smart enough” without proper validation, we begin to create fragile systems. Overconfidence in automation, aka: we get lazy. However, if the agent is just going to fix itself, do we care? (<– see, overconfidence is too easy) Context switching for the human brain is challenging. Even though less context is needed when managing agents (I don’t need to load the entire code base into my working memory), managing a dozen agents across a dozen apps, with varying technologies, will require next-level context management for us, the humans. Erosion of foundational skills can happen when we rely too heavily on AI, making us dumb. It may weaken our judgment and code quality instincts. There are some early studies out there in agreement. I would argue: if this is done right, this frees us to do more creative things, stretching new parts of our brains. But I’m an optimist. early studies out there Loss of system and architectural consistency. Today, agents run great locally, can optimize code within their immediate view, but frequently miss systemic design principles and the “bigger picture”. As context windows get larger and models get smarter, will this become a non-issue? If we implicitly trust AI and assume agents are “smart enough” without proper validation, we begin to create fragile systems. Overconfidence in automation, aka: we get lazy. However, if the agent is just going to fix itself, do we care? (<– see, overconfidence is too easy) Context switching for the human brain is challenging. Even though less context is needed when managing agents (I don’t need to load the entire code base into my working memory), managing a dozen agents across a dozen apps, with varying technologies, will require next-level context management for us, the humans. Your experience is needed… With the rise of Agent Directors, experience becomes all the more important. These roles prioritize oversight, coordination, and judgment over line-by-line coding. Unless… it won’t. The next iteration of AI tools could blow our minds; engineering skills and expertise will die like needing an ops team to manually push code into production. Maybe we, agent directors, are needed less as AI improves and can carry its own. I’m not sold that mind-blowing AI is coming too soon. AI, as we know it, is trained, spitting out best-chance tokens in order. Thinking models (don’t really think) can review, critique, and improve the chances, but it’s still a trained model. There is no creativity. No new ideas. No ingenuity. Human engineers have to remain on top of it to keep our code and the products we create optimal, impactful, and purposeful. Thinking models (don’t really think) You can’t prompt your way to being a senior engineer As AI becomes more embedded in engineering workflows, there’s a dangerous temptation emerging: we can shortcut real experience. We don’t need to train junior engineers anymore because AI can generate the code. We can downsize teams and rely on automation to fill the gaps. Prompting is enough. Lies. “Laying off developers because of AI is the dumbest move in tech right now.” (Perazzo, 2024) “Laying off developers because of AI is the dumbest move in tech right now.” (Perazzo, 2024) Perazzo, 2024 Organizations that take this path risk gutting their engineering pipelines. They risk losing the deep, critical knowledge that only comes from years of trial, error, and mentorship. And worst of all, they might not realize the damage until it’s too late. Systems start to rot, architectural coherence slips, and no one’s left who understands how it all works. We need to continue to invest in growing engineers, real ones. The kind who think, question, debug, and design. The kind who can lead in a future where AI is everywhere, but judgment still matters most. AI can write it. They decide whether it’s right. AI can write it. They decide whether it’s right. What might this look like? Intentional Growth for Junior Engineers Today’s junior developers need to be more intentional than ever about their growth. It’s incredibly easy to get lazy and let AI do the heavy lifting, but that short-circuits real learning. This isn’t a jr engineer problem, it’s a human problem. Those of us who came up before AI didn’t have to think so hard about being deliberate; we had to troubleshoot, write from scratch, and grind through the process. How many nights I burned chasing down bugs… That’s how the craft was built. Now, that same experience must be pursued with great intentionality. This isn’t a jr engineer problem, it’s a human problem. had had “AI can assist with code generation, but without mentorship, [our] code quality, security, and architecture suffer.” (Osmani, 2024) “AI can assist with code generation, but without mentorship, [our] code quality, security, and architecture suffer.” (Osmani, 2024) Osmani, 2024 To ensure today’s junior engineers grow into the thoughtful, system-oriented senior engineers we’ll need tomorrow, I encourage some practical techniques: Engage with AI AI isn’t going anywhere; it’s too hard to avoid. However, it doesn’t mean you have to give up and let it do everything, not allowing you to learn. It can be of great use and impact to your learning journey! Shadow the AI. Use AI to generate solutions and code, but take time to understand and question its decisions. Rewrite its code manually to compare approaches. Search the web for concepts it’s using and ensure they’re accurate. Treat the AI like a senior engineer, watch, question, and learn. Shadow the AI “Learning is now on-demand. I don’t need to know Python to build a Hello World app. I can build the app, then understand how it works.” (Palmer, 2025) “Learning is now on-demand. I don’t need to know Python to build a Hello World app. I can build the app, then understand how it works.” (Palmer, 2025) need Hello World Palmer, 2025 Question the AI. Ask the AI the same question more than once! If you get different answers each time, review what makes them different. Try different LLMs entirely, i.e. OpenAI vs Claude, and compare. Question the AI. Keep an AI contrast journal. Log the AI-provided solutions and comparisons alongside alternatives you’ve researched or come up with. Reflect on which choices are better and why. Keep an AI contrast journal Be critical of AI, don’t just assume it’s right (rarely is it that good). Ask questions, make changes, and engage with its outputs. Code without AI You don’t have to use AI. Give it a go! have Go old school (are the kids calling it vintage now?). Grab a coding book. Use Google to search, ignore its Gemini response, read blogs and forums, find code snippets, copy/paste, and write your own code! Go old school (are the kids calling it vintage now?). ignore its Gemini response AI-free problem-solving. Beyond coding, don’t use AI in debugging issues either. Nothing feels better than spending 4 hours troubleshooting a bug when you realize you mistyped a variable. AI-free problem-solving. Own a real component. Maintain responsibility for a small but non-trivial piece of software. Handle its evolution, debugging, and design choices. Own a real component. This is how our experienced engineers became experienced, building critical thinking and system design skills: hands-on code, learning new patterns through trial and error, and mind-numbing troubleshooting. Engage with People Outside of your own learning path, engage with experts who have gone before you. Read great code. Study well-maintained open source projects to internalize best practices and decision patterns. Use AI to explain certain complexities and ask why the authors might’ve chosen the direction they did. Read great code. Seek out human pairing. Participate in reviews with senior engineers, not just for code validation but to understand architectural and strategic thinking. Be curious, ask questions. Seek out human pairing. Share what you learn. Explain your work to others, how you reasoned through a complex situation. Consider speaking at user groups or blogging. Repeating what you learned to others reinforces these facts for yourself. Share what you learn. Your future is in your hands This takes time and effort, and chances are, it will happen outside of work. Make coding your new hobby, take on side projects, and be eager to learn! We don’t know the true future of software engineering, even if it is 100% AI-written code in 5 years. What you can learn today, through the above tips, will be invaluable. We’re going to need you, start today. Engineering leaders, we need to show up I appreciate it’s one thing to say all of this, and it’s another to enable some to do it. Unfortunately, most engineering teams are driven by deadlines and tight budgets. Letting an engineer learn costs time and money. Is the investment worth it? Yes, I think so! Organizations must invest now in developing the next generation of “engineers”, or Agent Directors, or whatever clever name someone comes up with. It’s the future, but we need to plan now. Failing to build future senior, experienced engineers introduces long-term gaps in leadership, technical expertise, and innovation. now
