If AI-Native Products Take Hold, the Pipes Lose First

Jiang Wu · 2024-02-18

*Originally published in Chinese on February 18, 2024.*

Hardware such as Rabbit's R1, a handheld AI device unveiled in January 2024, raises a question: if AI-native products take hold, who loses?

Who loses under an AI-native trend

Pipes weaken or disappear

Hardware is how users connect to the network. People cannot connect directly yet, not until brain-computer interfaces mature, so a carrier for services is needed (see my answer on the R1). If things go that way, the ones who lose are the pipes: products that gather services for users but do not carry the service themselves.

Services can be divided more concretely into physical and virtual. Physical services are easy to picture: everything outside that supports daily life, from food and clothing to housing and transport. Virtual services are broader, such as social, emotional, film and TV, entertainment, and games. By this logic we should ask who today carries the pipe and who actually provides the service. In practice, because of how things developed, most vendors of virtual services provide both the pipe and the service. Games, film and entertainment platforms are also where content gets produced and created.

Back to the physical side: some platforms today provide both service and pipe, and some provide only the pipe. If this direction does arrive, the logic of entry points and traffic will be redefined and redistributed. How users perceive getting a service, and how providers perceive the service itself, will have to be found and defined again.

The deciding factor is what form of delivery users ask for

At the core, a future hardware carrier only needs to provide a network connection. Setting aside bigger screens and better interaction, connecting does not demand much from the device. Interaction has changed, and what it needs is not something stronger or more complete but something closer to how humans, or any species, naturally act. That depends on upgrading the hardware, which takes thought from more directions and changes to the device system, whether it is a band, a watch, smart glasses, a head-mounted device, or even contact lenses. The AI-native shift does not depend heavily on changes in external devices. It depends more on how the cloud systems connect with each other and how they are understood and interacted with. What remains to be seen is who ends up carrying the delivery.

I lean toward saying that user demand for the form of delivery will decide whether we end up going AI-native at all. That has to be split by scenario and by what the user needs delivered.

How China's mobile internet shapes AI-native development

Delivery forms and the user's entry point

Suppose the user's hardware is the carrier that connects to the network, the entry point to services. Through what form the service is finally delivered may later matter little. Service providers will look at where the user is and offer an entry point in that form.

Pushed to the extreme, if you split things by this delivery logic, the party that finally interacts with the user will not care what form it takes to keep up with the change. It depends on the scenario in which the user needs the service, and the provider only has to distribute its front-end entry well. So how we define the result matters little. What matters is who can make their own platform or system the real entry point to the service.

The gap in core capability and the user experience

Our gap today is not one of form. It is a gap in the capability at the core. Whatever the form, users will in the end choose what fits their habits and gives the shortest path to the best channel for getting a service. But because of the gap in core capability, we cannot give users the best experience and feedback under equal conditions. We may even keep the standalone-app model we have now, because if the experience cannot meet user expectations, it is better not to renew and to deliver in a stable form instead. That is what the best business logic looks like right now, and it will exist for a while.

The worst result is a generational gap. Others are already using iPhones and we are still on landline phones. None of us wants to become North Korea, but we could only be North Korea.

Home · All articles · Topics · 简体中文