Agentic AI gives us something we never had before: instant feedback on every decision, every prompt, every output. Direct suggestions. A constant stream of next steps. That's efficient. That's productive. But what does that steady flow of reward do to our attention and behavior? Are we quietly becoming addicted to AI dopamine?

This lightning talk examines agentic engineering through the lens of dopamine, reward, and habit formation. When feedback is immediate and output quality keeps varying, a hard-to-ignore pattern emerges: prompt, evaluate, retry, one more time. That's exactly what makes this way of working not just powerful, but behaviorally addictive at the micro level.

The session breaks down how that reward loop operates in everyday engineering work. Why does "one more prompt" feel so logical? And why is stopping often harder than you'd expect? You'll walk away with a clearer picture of why you fall into this loop, how to recognize it in your own AI use, and what you can do to step out of it without losing the benefits AI brings.