The Hype vs. The Reality
The tech world is buzzing with talk of 'AI agents'—smart assistants that can book rides, order food, and plan trips on your behalf. Yet, a glance at the numbers from the giants of the gig economy tells a surprisingly different story. During its busiest
week in October, Uber reported that out of 320 million trips and deliveries, AI agents were used to book just four rides and about 500 Uber Eats orders. Similarly, DoorDash has seen only a few hundred orders placed via agents since the feature became available. These figures are a stark contrast to the industry hype, suggesting that for all the excitement, user adoption for these new tools is currently negligible.
The Speed 'Problem' That Isn't a Problem
The core question the headline raises is a valid one: why use an AI agent when the existing process is already lightning-fast? Uber's own CEO, Dara Khosrowshahi, pointed out that booking a ride on the app takes a median time of just 50 seconds, while reordering a favorite meal on Uber Eats takes about 90 seconds. For many users, delegating these quick tasks to a conversational AI might actually take more time and effort than simply tapping a few buttons. The current user interface on these apps has been refined over a decade to minimize friction. For simple, high-frequency tasks, it is not clear that an AI agent offers a meaningful improvement in speed or convenience.
DoorDash's Bet on Conversational Commerce
Despite the slow start, companies are still experimenting. DoorDash, for instance, launched a text-to-order AI agent that allows users to place orders via Apple Messages without ever opening the app. The idea is to make ordering as simple as texting a friend. You can ask for 'your usual' or request recommendations, and the agent handles the cart and checkout process. While initial uptake is low, DoorDash is betting that this conversational approach could become a new paradigm, especially for complex group orders or for users who prefer text-based interaction over navigating an app interface.
Airbnb's Vision for a Deeper AI Integration
Airbnb's approach to AI is perhaps the most different. CEO Brian Chesky has noted that trip planning is not a task most people want to complete as quickly as possible; it is often a fun, visual, and social process. Consequently, external AI agents have seen 'very non-material' use for bookings. Instead of focusing on speed, Airbnb is building its own proprietary AI agent, set to roll out in 2027, that will act as a comprehensive trip planner. This tool will generate a custom, visual interface with maps, photos, and comparisons, moving beyond simple text chat to create a richer planning experience that an external agent couldn't replicate.
The Real Goal: Beyond Customer-Facing Speed
If not to save customers a few seconds, what is the real purpose of this AI investment? The answer lies behind the scenes. For a company like Uber, AI agents are already having a massive internal impact. The company runs tens of thousands of AI agent executions weekly, with tools that autonomously write and fix code, manage internal services, and optimize engineering workflows. These 'Minions,' as Uber calls them, are boosting productivity by handling repetitive tasks, allowing human engineers to focus on more creative work. Similarly, AI is being used for hyper-personalized marketing, dynamic pricing, and sophisticated customer support that can proactively resolve issues like a canceled flight before the customer even notices. This is where the true value is currently being found.
















