What if chatbots weren’t always the right choice? In the GenAI era, they’ve become the default solution for integrating AI into user interfaces. Sometimes they work well, especially in purely conversational contexts, but too often they’re adopted automatically, without considering more effective alternatives. The result? Experiences that force users into tedious text-based interactions with AI, even when there are faster and more intuitive ways to get things done.

In this talk, we’ll explore when chatbots are truly the best option—and when they’re just a lazy design shortcut that makes interactions more cumbersome than they need to be. By revisiting the jobs to be done, we’ll dive into alternative interaction patterns that integrate AI in a more seamless and purposeful way. We’ll look at real-world examples of interfaces that mask LLMs behind guided selections, blend structured inputs with intelligent automation, or leverage hybrid models to create more efficient user experiences without forcing artificial conversations.

Because real innovation isn’t about making AI talk—it’s about making it work for us, in the right way, while respecting users’ time and needs.