Managing Drama Between Humans and Machines

AI is no longer a tool—it’s a collaborator. And like any collaborator, it creates friction.

Developers are frustrated with AI: it hallucinates, overreaches, and confidently gets things wrong. Meanwhile, AI is “frustrated” too—fed vague prompts, shifting requirements, and inconsistent human feedback, it produces brittle code, noisy tests, and unpredictable results. Add a layer of mutual fear—humans unsure they can trust AI, AI systems reflecting that uncertainty back—and the relationship breaks down fast.

I’ll speak candidly, with some humility: I’ve struggled with these dynamics for years. And in that time, one thing has become clear—this human–AI tension is now the single biggest drag on productivity and software quality.

This talk treats AI development for what it’s becoming: a counseling problem between humans and machines.

We’ll break down the real sources of conflict—misaligned expectations, lack of trust, poor communication (prompts), and missing feedback loops—and show how to fix them. You’ll learn how to set boundaries for AI, orchestrate instead of argue, and turn chaotic outputs into reliable systems.

Because the future isn’t humans vs. AI—it’s humans managing AI at scale. And right now, most teams are doing it poorly.