AI model companies have been stuffing in as many features as possible into their opening screens — ChatGPT users, for instance, can currently ask for responses, generate images, and do deep research from the same interface — but this isn’t necessarily how things could turn out in the long run.
Airbnb CEO Brian Chesky has weighed in on the future of AI interfaces, offering a counterpoint to the prevailing notion of a single, all-encompassing AI experience. Chesky’s perspective, delivered with a touch of historical context, challenges the idea of a singular “chatbot to rule them all.”
“There are some AI maximalist views that there’s going to be this one interface to rule them all, it’s going to be one chatbot, maybe ChatGPT. (These views say that there’s) going to be one interface, maybe voice,” Chesky said.
Chesky continued, arguing that the idea of a singular interface oversimplifies the future: “I think the problem with all those mental models is they try to simplify the future. Do we think the future is going to be simpler because of this new technology or more complex? I think the future is going to be more complex,” he said.
He then explained that he doesn’t believe in a singular entity that’ll dominate AI use: “I don’t think there’s going to be one chatbot. I don’t think there’s going to be one app. I don’t think there’s going to be one device, and there’s not going to be one interface.”
To illustrate his point, Chesky drew a parallel to the evolution of the smartphone: “Here’s an analogy: when Steve Jobs launched the iPhone during his keynote, he said the problem with the so-called smartphones of the era, with the bottom 40% of the phone being the keyboard, was that the keyboard didn’t change. The buttons didn’t change for each interface. Every application wants a different interface, so there shouldn’t, I don’t think, be one interface, one device, one way to do everything.”
“We’re going to have a lot more tools in the future, so it’s not this one thing, it’s going to be many things. That’s my view, just like there’s many tools,” Chesky said.
Chesky’s perspective highlights a crucial point: the potential for specialization and diversification in AI applications. While the allure of a single, unified AI interface is strong, the reality might be a more fragmented landscape. Instead of a “one-size-fits-all” solution, we could see a proliferation of AI tools tailored to specific tasks and industries. And this should bring hope to AI entrepreneurs who’re thinking of building their own products — there might be lots of need for lots of specialization and customization to products and experiences that the big AI labs won’t necessarily be able to provide.