What Do You Do With Your Nerds When AI Changes the Rules?
AI is reshaping technical teams—and leadership has to change too.
For decades, technical organizations had a quiet deal with their engineers. If you were good enough technically, you could mostly stay in the world of logic. We all know the archetypes: the brilliant but socially awkward developer, the quant who hates meetings, the engineer who only wants Jira tickets. This worked because technical work was scarce, and the ability to translate human ambiguity into machine certainty was a rare and highly valued skill.
But that deal is breaking down. The arrival of generative AI is fundamentally altering the nature of technical work, and it is doing so in a way that directly challenges the traditional refuge of the analytical mind.
AI Changes the Nature of Technical Work
AI does two things simultaneously: it makes technical production easier, and it makes interpretation and framing harder. The bottleneck in software development and data science is moving rapidly. It is no longer about writing the code itself. Instead, the bottleneck has shifted to problem definition, system design, and evaluation.
When an AI agent can generate a working component, its tests, and a deployment configuration from a well-scoped prompt, the sheer volume of code an individual can produce skyrockets. But this acceleration exposes a new constraint. As Michael Novati recently observed, the real bottleneck in the AI era is human. It is the coordination inside organizations, the alignment of incentives, and the ability to clearly articulate what needs to be built in the first place.
This shift means that technical work now requires significantly more human coordination. The very people who entered technical fields to avoid the messy, ambiguous world of human interaction are now finding that their jobs require them to navigate it constantly.
The Leadership Problem
Now leaders face a difficult question: what do we do with people who entered technical fields precisely to avoid this kind of work?
Organizations are currently experimenting with three possible responses. The first is to simply replace them. This is the narrative of layoffs driven by AI productivity gains. The problem with this approach is that it destroys deep institutional knowledge. You might gain short-term efficiency, but you lose the people who actually understand how your systems work under the hood.
The second response is to force them to become extroverts. Suddenly, every engineer is expected to present, coordinate, and lead meetings. The problem here is equally severe: you lose people who are brilliant but wired differently. You alienate the neurodivergent talent and the deep thinkers who thrive in focused, uninterrupted work.
The third response—and the only sustainable one—is to redesign technical organizations entirely.
Redesigning Technical Organizations
This is the interesting path. Instead of flattening roles and expecting everyone to be a generalist communicator, forward-thinking organizations are creating new structures. They are introducing roles like technical translators, architect roles, AI system designers, and evaluation specialists.
Not everyone has to become a communicator. But the interface between humans and machines must be owned. As Priyanka Vergadia points out, the old model of rigid, specialized silos is giving way to more fluid, cross-functional cells. In these new structures, you need both “M-shaped” engineers who can orchestrate across domains and “T-shaped” specialists who go deep into complex, non-promptable problems.
The Bottom Line
Strong organizations will protect their deep thinkers. They will pair them with translators and upgrade the system architecture around AI, rather than simply flattening roles and hoping for the best. This approach preserves cognitive diversity, which is more critical now than ever.
AI is not eliminating engineers. It is forcing organizations to learn how to work with them differently.
I’m Launching a Course!
So many AI projects die. And that’s not the fault of the tech nerds: They built the demo, and it worked. Still, 90% (yes, really) of all AI models never make it into production. So let’s dig deep into the big organizational underbellies, and let’s find out how we can make those numbers a bit better.
That’s the challenge I’ll be tackling in a new course starting April 21 at GenAI Academy, where we walk through how to actually move an agentic AI system from demo to production — including the organizational architecture required to make it work. This is for technical leaders, senior engineers, product managers, and AI/ML team leads.
I’m really excited to be able to bring what I’ve seen from the inside and outside to you in this format. You’ll experience me teaching live over 6 weeks! You’ll find all the details here: From Demo to Production.
Reads of the Week
AI Is Reshaping Engineering Orgs. Here’s How to Stay Ahead: In this piece for The Cloud Girl, Priyanka Vergadia argues that the traditional pyramid structure of engineering teams is being replaced by a “Cellular AI Org Model.” She explains how cross-functional, outcome-focused teams paired with autonomous agents are the future of technical work. This is essential reading for any leader trying to understand how to structure their teams for the AI era.
RDEL #99: How has AI impacted engineering leadership in 2025?: Lizzie Matusov breaks down the findings from the 2025 LeadDev Engineering Leadership Report. She highlights that while AI adoption is widespread, its transformative impact on productivity hasn’t fully materialized yet, requiring leaders to treat AI adoption as an organizational change rather than just a tooling choice. It’s a sobering look at the reality of AI integration in enterprise environments.
The Real Bottleneck in the AI Era Is Human: In this beautiful essay, Michael Novati explores why the massive acceleration in coding speed hasn’t translated to a proportional increase in shipped products. He argues that the true bottleneck is the human system surrounding production—coordination, trust, and regulation. This piece perfectly captures the tension between machine speed and human friction.