Airbnb has introduced a new AI-enhanced search feature as part of its autumn update, marking a pivotal moment for the hospitality giant as it navigates the evolving landscape of artificial intelligence. While the company has historically taken a measured approach to AI integration—CEO and co-founder Brian Chesky previously arguing that chatbots alone are insufficient for travel discovery—the latest rollout signals a shift in strategy. Chesky recently discussed these developments and his broader vision for the future of digital interfaces.
Chesky maintains his stance that traditional chatbots are ill-suited for e-commerce and browsing, primarily because they require cumbersome multi-turn conversations to deliver results. However, he acknowledges that Airbnb’s current search interface is not the final iteration. He envisions a hybrid model that balances the efficiency of conversational tools with the immersive, inspirational browsing experience that travelers crave.
“People actually derive more pleasure in planning and anticipating travel than they do traveling,” Chesky noted. “So you don’t want to take the inspiration and the browsing and the library experience of dreaming away.”
Beyond interface design, Chesky highlighted a critical social dimension: most current AI tools are designed for single users, whereas travel planning is often a collaborative effort. To address this, Airbnb is currently exploring “multiplayer” AI capabilities over the next three to six months, allowing multiple users to interact with the platform simultaneously.
The Case for an AI-Native Operating System
A central theme of Chesky’s argument is that the tech industry lacks a true operating system built specifically for AI agents. Currently, applications are layered onto existing platforms like iOS and Windows, which were not designed with agent-to-agent interoperability in mind. Chesky compared the current race to become the primary consumer AI agent to a situation where no one is building the foundational software developer kit (SDK) required to support it.
He cited ChatGPT’s early attempts at an app store as an example of what happens without a robust underlying OS and SDK. “If you want to be an app store like the iPhone, you need to have a software developer kit and an operating system like the App Store,” Chesky recalled telling Sam Altman. “Otherwise, none of our functionality is going to exist.”
Chesky suggested that if apps were to become truly agentic and interoperable, they would need to function like “apps with faces.” This could lead to a future where a macro Airbnb agent interacts seamlessly with agents from other services, potentially through standards like MCP (Model Context Protocol). However, he admitted that consumer AI has not yet been fully “cracked,” pointing out that using Airbnb through third-party agents like Muse or Instinct currently provides a suboptimal experience.
Voice and Internal Efficiency
Looking ahead, voice interaction is expected to play a significant role. Chesky predicted that humans will eventually speak to computers more than they type, with Airbnb planning to introduce voice agents for search and customer service this fall.
On the operational side, Chesky revealed that AI is already helping Airbnb ship features faster. Personally, he uses AI to bypass lengthy meeting chains, allowing him to gather information more efficiently. “AI is reducing communication layers,” he said. “One of my jobs as CEO is to try to get information. And the way I used to get information was having meetings. Now, I can just ask my AI.”
Leave a Reply