My first "real" job was being a programmer at a healthcare startup. I reported to a skilled programmer with far more experience than me, and I credit much of my foundational knowledge of software development to him. He was a great manager – he clearly communicated expectations, provided clarity without being overly prescriptive, and he gave me space to fail, iterate, and improve. He managed to have both soft skills and deep technical experience. In my experience, having a great manager is the single most important driving factor in my happiness at work.
Managing Humans
As engineers progress in their careers, they inevitably run into a fork in the road. Do they want to remain an individual contributor, or do they want to follow a management/leadership track? In my experience, engineers often struggle with this question, because the satisfaction of building something to solve a problem is what drew them to engineering in the first place. If they decide to become a manager, their technical experience and capabilities will take a back seat to the business of managing humans, which can be a scary proposition for an engineer. As a result, it's not surprising that the majority of engineers choose to remain individual contributors.
The business of managing humans is complex and requires a specific skillset. Empathy, clarity of purpose, selfless leadership, the ability to architect an organization of high performers – all of these are not core competencies required to be an engineer, with one key exception – clarity.Managing Robots
Engineers love precision. We have to – code is only worthwhile if it actually does the precise job that it is designed for. That means we often have to dig into requirements, ask pointed questions, and rapidly create the most clarity we can. The code is the easy part.
We now live in a world where agentic AI is being broadly adopted at every level of a business, but the place where it's gained the most traction is in software development. Large Language Models trained on hundreds of millions of lines of code are already deeply capable at building software, and are improving at a remarkable pace. Programmers all over the world are using Claude Code, Cursor, Kiro, and other agentic software development platforms to great effect, and a big part of their success can be directly attributed to programmers. Without them, agents are ineffective, inefficient technical debt machines.
An engineering team with a mature, spec-driven AI-DLC will require team members to ask the right questions, to tease out scope, and to create clarity that agents and humans can then use to write code. Engineers are the highest value assets in the chain–without them, agents will not have the right clarity and context to perform their job
So, guess what? Every individual contributor is now managing a team. It's just a team of robots. And the best news? These robot employees are delighted to do *precisely* what they are told.
Your People Make AI Work
If you've been reading my articles for a while, you know how passionate I am about responsible AI adoption that treats humans with the respect and importance that they deserve. AI is a tool for humans, not the other way around. Understanding your team's core strengths and the real, hype-free capabilities of AI will allow you to transform your team of individual contributors into a team of high-performing captains, precisely directing agents, writing and reviewing code alongside them, and going home at the end of each day fulfilled.
