Not everyone needs to become an AI engineer
There will be plenty of room for AI enabled engineers as well
In a world where engineers building software factories can deliver an unreasonable amount of code, it can be easy to think that all we need is more AI engineers who can build agentic harnesses. My guess is that in the future, we’ll see two distinct classes of engineering roles emerge.
AI Engineers - These are the folks that’ll be building the software factories. They’ll build and tend the orchestrators, they’ll review and curate the prompts, agents, skills and context. They’ll manage the memory systems and guardrails and ensure that your agentic software factory is well maintained and continually compounding.
AI Enabled Engineers - These will be the folks that will USE the software factories to deliver the features that power your business. They’ll understand good practices for working with a software factory, but they’ll spend most of their time focused on some combination of experiment design, specification and documentation of verifications. They’ll write the PRD that tells the factory what to build and define the acceptance criteria that tell it when it’s done.
And in a bit, we’ll all get lazy and just go back to calling them DevEx/Platform teams and feature/product (stream aligned) engineers like we do today.