Google accidentally showed a prototype of Jarvis AI capable of controlling computers
By ForGeeks Google has mistakenly released a preview version of its new AI prototype called Jarvis, which can perform actions on a computer without human intervention. The program appeared briefly in the Chrome Web Store, allowing users to see its functionality before access was denied.
The program was briefly available in the Chrome Web Store, allowing users to see its functionality before access was denied.
As reported The Information, a page on the Chrome Web Store described Jarvis as a “helpful assistant” that can perform simple tasks, from ordering groceries to booking tickets to searching for information. Jarvis works through a web browser, controlling your computer to perform everyday tasks automatically.
The Information reported.
Although the version of Jarvis that appeared in the Chrome Web Store was limited in features, the program showed its direction by fully automating certain tasks. When reporters tried to use the app, access to some features was blocked due to permission restrictions.
When reporters tried to use the app, they were unable to access some features.
Google quickly removed the page describing Jarvis, and the company plans to officially unveil the assistant in December. Google will enter into competition with other companies like Anthropic, which recently launched a public beta of its AI assistant Claude. Like Jarvis, Claude can also control a computer, enter text, activate buttons and move the cursor.
Anthropic’s AI assistant will also be able to control a computer, enter text, activate buttons and move the cursor.
