Enhancing Digital Interactions with AI Agents for Greater Convenience and Efficiency
Year: 2024
Type: article
Abstract: AI agents are software entities capable of performing digital tasks autonomously, adapting to new challenges and environments without direct human control. Recently, AI agents powered by large language models like GPT-4 have been integrated into operating systems such as Ubuntu, Windows, and macOS. These agents can install apps, edit images, remove tracking devices from websites, and automate routine digital tasks. OSWorld is a platform that supports the research of these agents. In this project, we reproduced the OSWorld environment and analyzed the agents’ ability to complete digital tasks using GPT-4o. The research results showed that AI agents can be used in a wide range of digital applications, offering significant practical benefits and assistance to users in various settings.
Authors: Mengtong Xiang, Ziyu Yao
Institution: George Mason University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7092781552
FWCI:
Cites:
Cited by:
Related to: