Revisited from the Foresaw archive. The argument and structure have been updated for this edition.
A new platform or AI tool can change the shape of an engineering role. It does not remove the need to understand a problem, reason about a system, or explain a tradeoff.
The most useful development plan combines technical depth with the habits that make that depth valuable to a team.
Learn through complete problems
Choose learning work that connects an idea to an observable result. Building, operating, and revising a small system can expose gaps that a tour of individual tools misses.
Follow the information across boundaries: from a user action to application behavior, storage, and operational visibility. That understanding makes unfamiliar technologies easier to assess.
Develop the judgment around the tool
Ask what a tool assumes, where it fails, and what simpler alternative exists. Learn to distinguish a convincing demonstration from evidence that something works under the required conditions.
AI-assisted work still needs someone who can recognize errors, test the result, and take responsibility for the change.
Make knowledge transferable
Write down decisions, explain the context behind a design, and help another person reproduce the work. Clear communication reduces the amount of knowledge that remains trapped with one specialist.
Give people room to grow
Organizations can support learning through review, pairing, focused experiments, and time to understand the systems they operate. New labels for roles are less useful than clear expectations and meaningful feedback.
A resilient team does not need everyone to know everything. It needs people who can go deep, learn responsibly, and work across the boundaries of their expertise.
Adapted from “The Future of IT Talent: Navigating Skills and Roles in the Digital Era.”
First published on Foresaw. Reworked for this edition.