Free tools Windows power users keep installed
One-click scans. No signup required.
Brian Chesky’s argument is that useful AI agents need infrastructure beneath the chat window: a platform layer and developer interfaces that let agents and apps exchange capabilities. That is not the same as Airbnb launching a phone or desktop operating system, and it is not an industry standard that already exists. In a TechCrunch interview published October 1, 2026, the Airbnb CEO also argued that travel discovery needs richer interfaces than a chatbot alone.
What does Chesky mean by an AI operating system?
In the interview with TechCrunch’s Ivan Mehta, Chesky describes an operating-system-like layer that would make AI capabilities available deeper in the software stack. Today, AI applications typically run on platforms such as iOS, macOS, or Windows; he does not consider those platforms AI operating systems simply because they host AI apps.
His idea is about coordination: agents need a way to access tools and services, and applications need interfaces that expose their capabilities to agents. Chesky says a complete platform would include a software-development kit (SDK) for that purpose. The goal would be for agents and apps to work together without every integration depending on a separate company-to-company deal.
He characterizes the current contest as a race to become the primary, or “quarterback,” agent. But a leading agent alone would not provide the underlying interfaces and controls he thinks the broader shift requires. He says: “It’s really up to Apple or Google, or somebody, to build a new platform for us to really make the true shift from apps to agents.”
#1 Best Overall
Why does he think chatbots are a poor fit for travel discovery?
Chesky’s criticism is not that chat is useless. It is that a chat transcript can be an awkward place to compare many options: it tends to reveal only a few at a time, and finding a useful result may take several turns. For a quick, narrowly specified task—his example is, “Book me a flight, I don’t want to look at it”—that may be exactly what a traveler wants. Airbnb trip discovery, he argues, often involves browsing, comparing, and building anticipation rather than simply delegating a transaction.
Travel also brings platform-specific tasks that do not fit neatly into a bare chat exchange. A traveler may need to browse listings, message a host, compare stays, verify identity, consult maps, or add other items. Chesky says a handoff to an agent—or a richer software-development interface—would need to preserve access to those capabilities rather than reduce the service to a text box.
For trips involving several people, he sees a need for “multiplayer” AI: a shared process that lets a group plan together. His interface position is therefore a mix of predictable, purpose-built controls and generative screens, not a choice between a conventional app and an all-purpose chatbot. He has previously put the point plainly: “I think that I’ve believed for a long time that a chatbot isn’t the right interface for e-commerce.”
Chat-first and browse-and-compose interfaces make different trade-offs
| Design question | Chat-first approach | Browse-and-compose approach |
|---|---|---|
| How many options are visible? | Usually a small number are presented in each response, so comparison may take more turns. | Can show multiple options together for scanning and comparison. |
| How does a person refine a choice? | By describing changes in successive prompts; the interaction can be conversational but may be lengthy. | By combining browsing with controls and generated assistance, allowing different kinds of refinement. |
| How does group planning work? | A standard one-person chat does not inherently provide a shared planning space. | A collaborative interface can make shared options and discussion part of the product; Chesky calls for “multiplayer” AI. |
| How predictable is the interaction? | Responses are generated, which can make the path less structured. | Designed interface elements can keep important actions and choices visible, alongside generative screens. |
| Can it complete platform-specific tasks? | Only if the agent has access to the relevant service capabilities and a way to invoke them. | Can retain native tasks such as host messaging, identity verification, maps, and adding items. |
This is a design comparison drawn from Chesky’s argument, not a published product test or a claim that one layout is best for every travel task.
Rank #3
- Incredibly Light. Surprisingly Thin. - LG gram is designed to go wherever you do. Weighing just 2.5 lbs. with an ultra-slim 0.7-inch profile, it slips easily into your bag and feels light in hand—making it effortless to carry, commute, and work from anywhere.
- Remarkably Light. Reliably Strong. - LG gram has passed seven military-grade durability tests, striking an impressive balance between a highly portable, lightweight metal build and the confidence to handle everyday movement and travel.
- Power That Last with Smart Efficiency - LG gram combines a high-capacity 72Wh battery with AI-driven power management to optimize efficiency based on your usage. The result is up to 32 hours of video playback for} long-lasting performance that keeps up with your day—at home, at work, or wherever you go.
- AMD Ryzen AI Performance - Powered by AMD’s AI-optimized Ryzen processor with Radeon Graphics and a built-in NPU, LG gram delivers smooth multitasking and responsive performance. Fast 32GB LPDDR5x memory and 1TB NVMe storage keep everything moving without slowdowns.
- Dual AI for Always-On Intelligence - LG gram’s Dual AI—powered by EXAONE 3.5, LG’s AI solution—combines gram chat On-Device AI and gram chat Cloud AI to deliver seamless assistance. gram chat On-Device AI enables fast document search and summarization directly on your PC, while gram chat Cloud AI expands capabilities when connected—so everyday tasks stay smooth, responsive, and uninterrupted.
How is Airbnb preparing for agents?
Chesky says Airbnb is making its infrastructure more agent-friendly. He imagines agents serving different areas of Airbnb and, eventually, a broader Airbnb agent that could interoperate with other agents through MCP. He also discusses voice agents. These are directions and expectations described in the interview, not evidence that a universal Airbnb agent or all of those integrations are already live.
The broader rationale is that agents might make services interoperable even when their companies have not built conventional direct integrations with each other. That possibility depends on services exposing usable capabilities and on agents being able to work with them safely. The interview describes Chesky’s vision; it does not demonstrate universal interoperability in practice.
Chesky also describes his own experience using the consumer agents Muse and Instinct to interact with Airbnb, saying Airbnb works poorly through them; he extends that criticism to hotel booking. This is his assessment, not an independent benchmark of those services. His conclusion is: “I don’t think we’ve cracked consumer AI.”
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What would an agent operating system have to manage?
Two 2026 arXiv preprints offer technical context for the idea, but neither establishes a settled design. The paper “Agent Operating Systems (AOS): Integrating Agentic Control Planes into, and Beyond, Traditional Operating Systems” describes how long-running agents that pursue goals, reason probabilistically, call tools, and adapt to feedback can strain conventional operating-system boundaries. It outlines possible system responsibilities:
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Best Value
- Scheduling: coordinating agent work and access to shared resources.
- Context and memory: managing information an agent needs across steps or over time.
- Tool and capability registries: making available actions discoverable and usable by authorized agents.
- Policy and trust enforcement: applying rules to what agents may access or do.
- Observability and audit: making actions visible enough to inspect and review.
These responsibilities highlight why the operating-system analogy goes beyond “an AI that can open apps.” A system must coordinate state, permissions, tools, and oversight as well as execute requests.
Is an AI-agent operating system already a standard?
No. The preprint “Towards an Agent Operating System – Lessons from Classical and Cloud OS” describes agentic systems as being in an experimentation phase, with many frameworks and protocols but no community consensus on core abstractions or guarantees. Its authors argue for precise, portable abstractions and standardization; that supports describing the field as unsettled, not treating one proposed architecture as accepted.
Different designs could place coordination in a user-space agent runtime, within a traditional operating system, or in a distributed control plane. They also have to make choices about how to preserve context and state, mediate tools and permissions, and observe and audit actions. The interview and the preprints frame these as open design questions rather than a consumer product bake-off with a winning approach.
That distinction matters: Chesky’s platform-layer thesis is an argument about what agents and apps may need to interoperate, not proof that such a universal layer has shipped or that Airbnb has built one. His related point about travel is equally specific: making services agent-accessible should not require stripping away the browsing, comparison, and collaboration that some travelers value.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
