I've been chewing on something after listening to Dan Shipper on Lenny's Podcast and Dharmesh Shah on My First Million. Both conversations touched on how the individual contributor role is fundamentally changing in the age of AI.
The End of the IC?
For decades the individual contributor (IC) has been the basic unit of knowledge work—writing code, drafting copy, analysing data one task at a time. Layers of management exist to coordinate that output, but the throughput of any organization is the work that can be done by the ICs.
Generative AI flips that equation. When a model can generate a first draft in seconds, productivity is no longer gated by how many hours a single person can type but by how many well‑directed agents they can run in parallel.
From Craft to Coordination
The next level of productivity belongs to people who can orchestrate fleets of AI tools. Their toolbox looks more managerial than technical:
- Delegation – translate fuzzy goals into prompts, guardrails and evaluation criteria.
- Taste – identify the one useful answer in a sea of plausible ones.
- Systems thinking – invest in repeatable pipelines instead of one‑off deliverables.
- Feedback loops – tune agents the way you coach teammates — continuously onboarding, training, directing.
In short, everyone becomes a manager, even when there's no one to add to the org chart.
The Hard Part: Making the Leap
Here's the challenge: we're already busy. Backlogs are full, calendars are packed. When can we step off the treadmill and build an army of agents that will do the work for us?
For me in my small software company, I play two roles, software developer and president.
In my role as software developer, I'm diving head first into AI driven workflows. Writing lines of code used to be the most time consuming and expensive part of building a software company, now it's one Claude Code (or Cursor, or Copilot, or Codex, or ...) prompt away. The bottlenecks have moved to defining what needs to be built and verifying it does what you want.
In my role as president, I want to do a better job of helping my team identify AI driven tools and processes that can save them time and move them along the IC to AI Orchestrator spectrum. Dan Shipper at Every has an internal role called Head of AI Operations who meets with everyone on the team individually and helps them build out those AI workflows. However it happens, it's clear to me that AI Orchestrators are the future, and the faster we get there, the more efficient we can be and the more value we can deliver to our clients.