For more than fifteen years, web browsing has relied on a well-established model: an
address bar ,
tabs as far as the eye can see, and
sites that the user mentally assembles to accomplish a task.
A familiar mechanism, but one increasingly challenged in the age of AI. While new players seek to position themselves,
such as Perplexity’s Comet or
OpenAI’s ChatGPT Atlas , Google has decided to respond. With
Disco , presented by its Labs division, the company explores a radically different vision of the web…
Contrary to what one might think,
Disco is not simply an evolution of Google Chrome . On its
introductory page , Google presents it as a still-experimental tool, based on Chromium, but profoundly transformed by AI. Here,
traditional navigation disappears in favor of a conversational interface, and the address bar vanishes, giving way to a continuous dialogue between the user and the browser.
With this project, Google no longer wants to display the web as is, but to interpret and restructure it according to the expressed intent. Disco acts
as an agent that explores, analyzes, and synthesizes information in the background. Furthermore, Google no longer speaks of navigation, but of human-computer interaction, in which the browser becomes an active participant.
The core of this approach is based on GenTabs , a technology that allows Disco to create custom web interfaces, generated on the fly by AI.
In practice, instead of opening a multitude of tabs to compare information, the user makes a request, and Gemini assembles a temporary application that centralizes the useful data.
A travel project, for example, can generate a single page that integrates this interactive map, along with transport schedules, weather, and booking options. This interface doesn’t exist as a traditional website, but is coded instantly by AI.
A promise that worries publishers
Behind this technical feat, Disco raises major questions for the web ecosystem. Even if Google assures that each generated element links back to its original sources and that the sites are indeed accessed in the background, if the generated interface fully meets the user’s needs, they might no longer need to visit those sources.
By redefining the browser as an information architect , Google is positioning itself even more as an intermediary between the user and content. This is an appealing development from a user experience perspective, but potentially destabilizing for publishers, whose visibility and revenue depend on direct access to pages.