Author: Stephen Katte, CoinTelegraph; Compiler: Tao Zhu, Golden Finance
Google's artificial intelligence research laboratory DeepMind said that its latest artificial intelligence model Gemini 2.0 will be an important tool for building more advanced artificial intelligence. The basis of agency.
Google DeepMind CEO Demis Hassabis and Chief Technology Officer Koray Kavukcuoglu said in a blog post on December 11 that the Gemini 2.0-powered artificial intelligence agent released on December 11 can understand complex Direct, plan, reason, take action across sites, and even assist with video game strategy.
Hassabis and Kavukcuoglu said: "Practical applications of artificial intelligence agents are an area of research full of exciting possibilities.
"We are exploring this new frontier through a series of prototypes , these prototypes help people complete tasks and get things done. ”
According to Hassabis and Kavukcuoglu, there are currently multiple experimental AI assistant projects powered by Gemini, each with different functions.
One of the projects is called Deep Research, which helps users explore complex topics by searching the web to create a multi-step research plan and then generating a long report on the results.
Project Astra is a general-purpose AI. Assistant, primarily geared towards daily tasks such as providing suggestions and advice based on user-provided prompts, such as how to do laundry or more information about landmarks
Project Mariner focuses on creating an AI agent that can control you. Chrome browser, move the cursor, click buttons, fill out forms and browse websites
According to Hassabis and Kavukcuoglu. The projects are "still in the early stages of development," but they hope to make them "widely available in future products" after testing and further development.
"That's not the case yet. Early on, but the Mariner project shows that it is technically possible to navigate in the browser, although currently it is not always accurate and slow to accomplish the task, but this will quickly improve over time. ”
Meanwhile, Project Jules is being developed as a developer assistant that can be integrated directly into GitHub workflows and help with tasks such as coding and planning.
Hassabis and Kavukcuoglu said they also use Gemini 2.0Agents built for video games can advise players on next steps in real-time conversations and search for "rich game knowledge" online.
"We are working with leading game developers such as Supercell to explore how these agents work and test their ability to interpret the rules and challenges of a variety of games, from strategy games to farming simulators," they said .
In November, Marc Benioff, CEO of US cloud computing software company Salesforce, said the future of artificial intelligence lies in autonomous agents, not large language models (LLMs).
“I actually think we’ve reached the ceiling for LL.M.s now,” he said on the WSJ’s “The Future of Everything” podcast on Nov. 23.
Nvidia is also focused on positioning itself at the forefront of trends.
“We’re seeing the number of AI-native companies continue to grow. Certainly, we’re starting to see enterprise adoption of agent AI really as the latest wave,” Nvidia CEO Jen-Hsun Huang said on the company’s third-quarter earnings call in November stated at the meeting.
In addition, Hassabis and Kavukcuoglu said the team is "experimenting with agents that can help in the physical world" through robotics. Currently, Google’s AI agents are only released to testers and developers.