Computers can be made to be so smart that it matches, or even surpass human intelligence.
But think about the possibilities if such intelligence is able to interact with what humans have interacted for years. The result could be astounding, and that is what tech companies are after. Many of them, big and small, are competing in an arms race, to see who reaches AI supremacy first.
And Google, one of the bigger players in the field, has a long list of AIs it can boast under its sleeves.
But pretty much nothing compares to 'Jarvis'.
This is because the AI is capable of using a computer, meaning that it should be able to do things no other AI commoners can ever do.
It all began when The Information reported that Google has a project called Project Jarvis, which it refers to as a "computer-using agent".
the store page describes the Jarvis prototype as “a helpful companion that surfs the web for you”
Google initially planned to unveil Jarvis in December alongside an upgraded Gemini, which is supposed to be the next generation of its large language models.
However, according to more recent reports, an early version of Jarvis was mistakenly uploaded to the Google Chrome Web Store, where it was briefly available as an extension.
It's said that the AI-powered tool is capable of taking control of a user’s computer to perform everyday web-based tasks.
The store page describes the Jarvis prototype as "a helpful companion that surfs the web for you."
According to the publication, this Jarvis can handle tasks, such as purchasing products, making reservations, and conducting online research.
In other words, the AI has the ability to take control of a computer to complete these simple tasks without manual human input.
However, users who was managed to download it discovered it wasn’t in full working mode to them due to permissions issues.
Google swiftly removed the extension from the store, suggesting that it was a leak, whether it was an intentional or not.
Regardless, the move signals Google’s competitive stride toward AI that can autonomously use computers, a feature similar to Anthropic’s Claude 3.5 Sonnet, released in late October.
Claude’s “computer use” functionality allows it to perform tasks as a human would—by viewing screens, moving cursors, clicking buttons, and typing. With this kind of technology, users could soon instruct an AI agent to buy concert tickets, reserve dining spots the moment they open, or search through files in seconds.
Jarvis is one of the names famously tied to the Marvel Cinematic Universe (MCU).
Written as J.A.R.V.I.S., it's a "Just a Rather Very Intelligent System" that is a AI created by Tony Stark, which later help controls, and also co-pilot his Iron Man armors. Over time, the AI that is continuously upgraded, is also tasked with running business for Stark Industrie,s as well as the security for Stark's Mansion and Stark Tower.
In all, it's a natural-language user interface computer system that is named after Edwin Jarvis, in honor of the butler who worked for Howard Stark and the Stark household.
Google Jarvis is far from doing any of what the MCU Jarvis can.
But the tech titan knows where it's going.