Google unveiled Project Mariner; A tool for web surfing with the help of artificial intelligence

Google recently and simultaneously with the introduction of the new version of Gemina artificial intelligence, also unveiled Project Mariner, a large part of which is based on the Jarvis assistant, which we have seen published in the past.

Artificial intelligence tools with agency to perform various tasks are one of the hot topics these days in the world of technology, and most companies are also trying to introduce new products in this direction for their users. Recently, during the introduction of Geminai chatbot version 2.0, Google has unveiled a tool to help users while surfing the web called Project Mariner, which is currently under development and only limited people will be able to use it.

The Mariner project is one of the products of Google’s DeepMind division, which has a long history of working on artificial intelligence tools. Of course, this part of Google has a not-so-bright history of stopping its projects, and it is by no means far-fetched that something like this will happen to Mariner in the future. However, the official announcement and limited preview for users show that Project Mariner is a bit more advanced than other DeepMind products, and we may see a public version in the future.

Among the capabilities of Google’s web browsing assistant is to perform user tasks automatically; So that it can take full control of the mouse pointer and keys, fill out online forms and move between different websites. Tools like Project Mariner can have a significant impact on how users interact with websites and even change the design of web pages in the near future. It is also possible that many online sellers of products or even content publishers who count on receiving real visitors from the Google search engine will have to change their former practice.

What are the capabilities of Project Mariner web browsing assistant?

In the published preview of the Mariner project, we see a window on the right side of the Chrome browser where the user can enter the desired commands. Of course, to activate this window, you must first install its special extension on the browser, which is currently available to some selected Google testers. Finding information from websites, ordering products or even registering on different platforms are among the commands that users can give to Google’s artificial intelligence assistant to take care of them.

In the preview of Project Mariner, it can be clearly seen that Google’s web browsing assistant is not very fast, and there is a gap of about five seconds between each movement of the mouse pointer by it. One of the main reasons for this issue is the need to record a screenshot from the screen and send it to Google servers for processing and providing a new command, which will take time. It is also not clear how, if similar tools become widespread, their processing needs will be met with a huge amount of recorded screenshots. The same issue causes another limitation, which includes the constant use of the monitor and browser; As a result, the user cannot assign tasks to the web browsing assistant and perform other tasks on the same window at the same time.

Google recently unveiled another new artificial intelligence-based feature called deep search, which allows users to perform advanced and multi-step searches and can provide people with multiple search results in the form of a summarized text.