Eseandre Mordi
⬤ Agentic AI hit a breakthrough in 2025, making task execution cheap enough to fundamentally change how work gets done. Every valuable project has three phases: picking the right problem, doing the actual work, and checking if it's any good. AI now dominates the middle part—the execution—because these systems learn action sequences from data and get better through trial and error.
⬤ Execution relies heavily on "tacit knowledge"—the stuff people know how to do but can't easily explain in a manual. Old-school automation couldn't bridge that gap. Today's agentic AI learns by actually doing things, which means it's cracking tasks that used to need human hands and human intuition across tons of different fields.
⬤ When execution becomes dirt cheap and unlimited, value shifts to what bookends it: asking smart questions upfront and judging results afterward. That's where judgment calls happen—prioritizing what matters, reading the room, making tradeoffs. Humans still own that territory.
⬤ Work is being redesigned around a new role: humans as "Chief Question Officers." You set the target, define what success looks like, and evaluate what comes back. Agentic AI isn't killing jobs—it's relocating them. People move from doing tasks to designing systems, supervising AI execution, and making sense of what it produces.
Eseandre Mordi
