Google’s 'Project Jarvis' AI may soon handle your flight bookings and shopping
The race to build fully autonomous AI agents seems to be heating up. Just last week, Anthropic introduced an updated Claude 3.5, allowing it to interact with computers in ways similar to human actions.
Now, Google is reportedly working on its own “computer-using agent," known as Project Jarvis, which aims to automate common web-based tasks like booking flights, doing online research, and finding shopping deals—all within Google Chrome.
Jarvis’s functionality reportedly relies on capturing regular screenshots to interpret the content on a screen before taking actions, such as clicking buttons or filling out text fields.
Due to its cloud-based setup, some sources note that Jarvis takes a few seconds between actions—likely allowing for the precision needed in handling complex, multi-step tasks.
Reports from The Information further suggest that Google’s Gemini 2.0 model will drive Jarvis, enhancing its capabilities to interpret and respond to complex commands more effectively.
To test and fine-tune these features, Google may initially release Jarvis to Gemini Advanced subscribers in a phased rollout, potentially beginning in December. This measured approach could allow Google to gather early feedback and address any issues before a wider release. However, these timelines remain flexible as development continues.
Sign up for Techloy
Techloy.com publishes information about companies, products, careers, and funding in the technology industry across emerging markets globally.
No spam. Unsubscribe anytime.
Project Jarvis also enters a competitive field where similar tools are emerging. Microsoft’s Copilot Vision and OpenAI's ChatGPT offer related functionalities, and Apple is rumored to be developing its own screen-aware AI assistant.
Yet, by integrating Jarvis directly within Chrome, Google could provide unique value to users relying heavily on browser-based applications. This approach also resembles Rabbit’s "large action model," which also seeks to streamline web interactions through automation but is yet to be fully available.
Should Jarvis progress as planned, it may redefine how Google Gemini's 330.9 million users approach repetitive online tasks, potentially enhancing productivity by automating everyday activities.
With the anticipated December preview approaching, further updates on Project Jarvis could indicate a shift in how AI integrates into daily digital workflows, shaping the future of web-based automation.