Google has taken a significant leap into the future of web interaction with the introduction of Mariner, an AI agent built on the backbone of its latest AI model, Gemini 2.0. This innovative system is designed to perform tasks on behalf of users directly within the Chrome browser, marking a new era in how we interact with the web.
Mariner leverages the advanced reasoning, planning, and memory capabilities of AI agents, allowing users to simply type their requests into their browser for the AI to execute. “We’re basically allowing users to type requests into their web browser, and have Mariner take actions,” explains Jaclyn Konzelmann, project manager at Google, in an interview with The New York Times. This is part of Google’s broader vision to embed AI agents into daily digital life, enhancing productivity and simplifying complex tasks.
Gemini 2.0 itself is a significant evolution from previous models, with a focus on what Google calls “agentic AI.” This term refers to AI that can not only process information across various modalities but also reason and act autonomously. “Gemini 2.0 represents a significant upgrade from its predecessor,” says Sundar Pichai, Google’s CEO, highlighting the model’s ability to think several steps ahead in task execution.
One of the standout features of Gemini 2.0 is its “Deep Research” capability, where Mariner can sift through information, analyze it, and produce detailed reports, a task usually requiring human-level reasoning and synthesis. This functionality aims to make Mariner more than just a tool but a partner in research and information gathering.
However, Mariner is not without its caveats. Currently available in a limited release, its deployment is cautious, with Google acknowledging that the technology is still in experimental phases. “Is [Mariner] always accurate? Not yet. It is still an experimental technology,” Konzelmann notes, emphasizing the need for human oversight in the loop to ensure accuracy and trust in the system.
Mariner can assist with tasks like online shopping by filling your basket or booking services, but final decision-making remains with the user, ensuring a balance between automation and human control. This approach not only enhances user experience but also responsibly integrates AI into daily tasks.
The introduction of Mariner and the broader capabilities of Gemini 2.0 have undoubtedly stirred the competitive waters, especially with rivals like OpenAI enhancing their own models like GPT-4 to keep pace. OpenAI’s focus on both consumer and enterprise applications through ChatGPT demonstrates the intense race to dominate AI integration in digital platforms.
As we look forward to 2025, when Mariner might become more widely available, Google’s current limited release is a tantalizing preview of how AI agents could transform our interaction with technology, making the web not just a place to browse but a space where AI works for us, under our supervision. This development not only sets a new standard for what’s possible with AI but also challenges other tech giants to innovate further in the realm of practical, everyday AI applications.
Your First 10 AI Skills
10 practical AI skills, copy-paste prompts and a 7-day plan to start using AI with confidence.
Download the guide →