Three Career Tracks for Developers in the Age of AI

For 60 years, the job of a developer was straightforward: a human writes code, a machine executes it. That model is ending. The fundamental unit of work is shifting from the instruction to the token, a unit of purchased intelligence. You describe what you want, feed the context, buy enough inference, and the machine figures out the implementation. That shift has split the software career into three distinct tracks.

Track 1: The Orchestrator

The orchestrator does not write code. They specify outcomes and manage the AI systems that produce those outcomes. Think of it as a factory manager for AI. The speaker cites Strong DM as an example: a three-person engineering team where nobody writes code and nobody reviews code. Agents write, test, and ship autonomously. The humans write specs and evaluate results.

This is the track that enables solopreneurs and small teams to operate at a scale previously requiring much larger organizations.

Track 2: The Systems Builder

Someone has to build the infrastructure that orchestrators use. The agent frameworks, the evaluation pipelines, the routing layers that send the right task to the right model at the right cost. This is a smaller track in terms of headcount, but the pay ceiling is high because the work is technically deep and the demand is growing fast.

Track 3: The Domain Translator

This is the track the speaker considers the biggest opportunity and the one nobody is talking about enough. Domain translators are people, often not traditional developers, who combine enough technical fluency to work with AI and deep knowledge of a specific industry. The dental practice specialist who can now build scheduling tools. The construction expert who can automate project timelines. The insurance compliance analyst who builds audit tools instead of just filling out spreadsheets.

These people understand which problems are actually worth solving in their domain, something a generalist engineer cannot easily replicate.

Who Is Most Exposed

The person most at risk is the one in the middle of the old distribution: someone who writes competent code but lacks either deep systems expertise or deep domain expertise. The value of generic code production is approaching zero. You need to pick a track because the middle ground of the old model is disappearing.

Key Takeaways

• The unit of developer work is shifting from writing instructions to purchasing inference (tokens)
• Orchestrators manage AI agents that write, test, and ship code without human coding
• Systems builders create the infrastructure (agent frameworks, eval pipelines, routing layers) that orchestrators depend on
• Domain translators combine technical fluency with deep industry knowledge to build tools for specific fields
• Generalist developers without deep systems or domain expertise face the most displacement

Resources

• Strong DM -- Referenced as an example of a team where agents handle all coding autonomously