AI Seminar 139: Anita Sarma

Title: Intentional Evolution: Three Ways to Evolve and One That's a Trap

Abstract: 

It starts as the illusion of fluency. AI hands back work that reads clean and feels finished, and that polish passes for your own competence. Some delegation is healthy. The trap is handing off the part of the work that was your craft once. Do that, and the skill stops evolving, so the next task leaves you more dependent than the last. We call it the cognitive debt cycle, and it compounds: every handoff feels like progress while your ability to work without help erodes.

Open source feels this first, where a flood of machine-written contributions meets maintainers whose attention is fixed and exhaustible. This talk draws on what we have heard from thousands of developers, students, and open-source contributors. What counts as meaningful work? What happens to our professional identity and skills when the craft gets handed off? What holds a community together when we turn to bots? And what about the people just starting out?