The built-in functions of the different web browsers we work with are becoming more and more advanced. This is the case of the most widely used proposal, Google Chrome, which is now testing a more efficient advanced translation method.
One of the biggest handicaps we face when moving around the Internet is the language we encounter on the different websites. Not everyone is fluent in English, not to mention other less common languages. That is precisely why Google is currently testing a new API that uses AI machine learning models.
All of this is done in order to offer real-time language translation of the entered text and facilitate the translation of web pages. Let’s say that this is a feature developed by an AI team for Chrome. Its main goal is to add translation functionality to the web browser and the ability to download additional language models to translate texts, all in real time.
It’s true that Google Chrome and Edge already have built-in translation features. However, they sometimes have trouble translating web pages with dynamic or complicated content. For example, Chrome is often unable to correctly translate all sections of an interactive website.
In these cases, websites have their own translation tool, but this can be resource-intensive and slow compared to the browser’s built-in feature. Google plans to offer developers access to Chrome’s upcoming AI-powered translation feature via a JavaScript API.
How AI helps Chrome translate websites
It is worth mentioning that at the moment, to carry out translation in these more complex cases, websites have to resort to cloud APIs. They can also use their own translation models and run them using technologies such as WebAssembly and WebGPU. But Google wants to go one step further.
The new feature introduces a new JavaScript API to add a browser’s language translation capabilities to web pages. This can serve as a simpler, less resource-intensive alternative to what is currently the case. The new API simplifies this process by using Chrome’s built-in translation engine for existing content or input text, aided by AI.
The feature actually downloads a machine learning model to correctly translate text if it is not built into the browser. However, the proposal, in its current form, raises some privacy concerns. This could allow a website to fingerprint a user based on the languages ​​supported by the browser. However, the feature is still in the testing stage and still has a lot of room for improvement.
All of this before it reaches the final, stable version of Google Chrome to use AI and speed up the translation of complex web pages. Right now, Google is exploring and working on the feature to also share it with the rest of the Chromium project.
