Are You ‘Agentic’ Enough for the AI Era?

A viral discussion spearheaded by Notion's leadership highlights a seismic shift in tech talent valuation, where 'agency'—the ability to direct AI agents—is surpassing traditional coding skill as the premium human capability. Notion COO Akshay Kothari and co-founder Simon Last argue that with AI coding agents like Claude Code and Codex automating execution, human value now lies almost entirely in deciding what problems to solve and managing the AI workforce. This perspective, echoed by investors like a16z's Jennifer Li who see AI tool ignorance as a 'red flag,' suggests a new hiring paradigm where a few highly 'agentic' individuals who can orchestrate AI are worth 'thousands of engineers.'

The technical reality underpinning this shift is the widespread adoption of AI coding assistants, with some companies mandating their use. Simon Last, who has stopped writing code manually after two decades, exemplifies the extreme end of this trend, running up to four AI agents concurrently and experiencing 'token anxiety' if they aren't working. This creates a new skillset centered on delegation, review, and high-level direction, leading to what Notion claims is higher product velocity. The implication is that this 'agentic' model will soon expand beyond software into finance, legal, and creative fields, forcing professionals everywhere to assess their capacity to manage AI, not just perform tasks.