Google is preparing an experiment with built-in artificial intelligence in the Chrome browser. The company invited several people to participate in a closed test of integrating a large machine learning language model into Chrome.
To access the AI model from web applications and browser extensions, the company offers the Prompt API, which allows you to send requests in natural language, similar to chatbots.
It is assumed that a large language model built into the browser will simplify the execution of AI tasks in web applications and eliminate the need to worry about installing and managing language models.
The model will automatically use the GPUs and NPUs available on the system to speed up the work, or simply use the CPU. The advantages of running the model on the user's system include maintaining the confidentiality of processed data, the ability to continue working in offline mode in the absence of a network connection or if there are problems with the quality of communication, reducing delays when sending requests, and eliminating dependence on external services.
The Prompt API can process and sort data, taking into account the context of previous queries and data within the session. In addition, it can be used to identify optimal options, such as selecting an appropriate icon from a list of emojis for a particular comment on the site.
Google said the API is currently under active development and will be expanded and changed based on user feedback and preferences before its final version is approved.
In the future, Google will provide an Origin Trials plugin, which will provide the ability to work with experimental APIs from applications running from localhost or 127.0.0.1 after registering and receiving a special token that will be valid for a limited time for a specific site.
At the same time, Google is working with browser manufacturers to standardize the APIs being developed.
