Editorial illustration for Developers' Value Shifts to Directing AI Coding Agents, Not Writing Code
AI Coding Agents Reshape Developer Roles and Skills
Developers' Value Shifts to Directing AI Coding Agents, Not Writing Code
Why does this matter now? Because the daily grind of typing loops, debugging, and refactoring is being replaced by a new kind of oversight. While the tech is impressive—large‑language models can spin up functional snippets in seconds—many developers find their core contribution slipping from hand‑crafting code to issuing high‑level commands.
The transition feels odd to those who built careers on the tactile act of writing, yet it also opens a niche for people who excel at framing problems for machines. Companies that once measured productivity in lines of code are redefining metrics around prompt design and task allocation. The result?
A growing segment of the workforce is measuring worth by how well they can steer AI coding agents, not by the code they type themselves. Simon Last, a Notion co‑founder, observes that this shift is unsettling for some but presents an opportunity for the industry’s most “agentic” talent.
For many, a massive piece of their value is now tied to deciding what AI coding agents should work on. The shift is unsettling to those who enjoyed the act of sitting down and actually writing code, but for some of the industry's most "agentic," it's an opportunity. Simon Last, another Notion cofounder, uses AI coding agents more than anyone I've ever met.
He coded for nearly two decades--then abruptly stopped. Now, he's using up to four AI coding agents at any given time (he prefers Codex to Claude Code).
Is the developer’s role now about steering machines rather than typing? For many engineers, value has migrated from the act of writing code to the judgment of which tasks to assign to AI agents. The shift feels unsettling to those who once found satisfaction in the craft of coding, yet Notion’s co‑founders Akshay Kothari and Simon Last see it as a chance to exercise a different kind of agency.
“Today’s agents might already be more capable than all three of us here in the room,” Kothari remarks, underscoring how quickly the balance of skill is changing. Because the tools are improving, the industry is rewarding the ability to frame problems for the AI rather than solve them manually. Whether this redefinition will sustain long‑term career relevance remains unclear; the article notes that the transition is still unsettling for a sizable portion of the workforce.
In practice, developers are being asked to become directors of code‑producing bots, a role that demands strategic thinking but also raises questions about the future of hands‑on programming.
Further Reading
Related Reading
Business & Startups OpenAI, a Series F San Francisco startup founded in 2015 by eight pioneers
Business & Startups Terminal-Bench 2.0 launches with Harbor, testing any container-installable agent
Business & Startups Zuckerberg Unveils Meta Compute to Build Global AI Infrastructure
From Archives OpenAI expands London office, citing UK talent and research hubs
From Archives Experts warn against padding AI‑screened resumes with irrelevant data
Common Questions Answered
How are AI coding agents changing the role of software developers?
AI coding agents are shifting developers' primary value from directly writing code to directing and framing problems for machine learning systems. Instead of manually typing loops and debugging, developers now focus on high-level strategic guidance and determining which tasks AI agents should tackle.
What approach does Simon Last take with AI coding agents?
Simon Last, a Notion co-founder, uses up to four AI coding agents simultaneously and has completely stopped writing code himself after nearly two decades of traditional programming. He represents a new breed of developer who leverages AI agents as primary tools for software development.
Why do some developers find the transition to AI coding agents unsettling?
Many developers have built careers around the tactile and creative process of hand-crafting code, finding satisfaction in writing software directly. The shift to primarily directing AI agents feels disruptive to their professional identity and challenges their traditional understanding of software development work.