On March 5, 2025, the Chinese tech startup Monica launched Manus, an advanced AI agent capable of performing real-world tasks autonomously, marking significant progress for China on the global AI stage. Unlike traditional chatbot models, Manus can think, plan, and execute tasks without requiring continuous human oversight, showcasing features never before seen at such scale.
Positioned as the world's first fully autonomous AI agent, Manus operates independently within the cloud, which allows it to continue tasks even when users disconnect. This unique capability sets it apart from existing models like ChatGPT and Google's Gemini, which rely heavily on user interaction. According to Yichao Ji, Manus' co-founder and chief scientist, the AI can accept prompts for complex tasks such as creating websites, planning trips, and analyzing stocks—all executed through streamlined, multi-step processes.
Since its official debut, Manus has garnered worldwide attention. It reportedly surpassed OpenAI's DeepResearch on the GAIA benchmark—an important metric for assessing AI performance. This milestone speaks volumes about the rapid technological advancements made by Chinese firms; particularly Monica, part of the Butterfly Effect group, which stands at the forefront of these developments.
Users have likened their interactions with Manus to those they have with ChatGPT, providing simple prompts to initiate complex analyses or creations. For example, if prompted to “Create a 7-day travel itinerary to Bali within a budget,” Manus autonomously scours the web and compiles detailed itineraries, complete with links, maps, and travel recommendations. This functionality highlights its potential to revolutionize productivity.
Manus' advanced capabilities include producing comprehensive reports, interactive presentations, and data-driven outputs such as visualizations and spreadsheets. The system's intelligence enables it to anticipate problems, optimize solutions, and manage several tasks concurrently, offering unprecedented efficiency.
Yet, not without challenges, early tests experienced hiccups, including factual inaccuracies and misattributions when citing sources. Critics have raised concerns about the reliability and accountability of using such autonomous systems. Should something go wrong, questions arise: Who is liable when AI makes errors?
The global economic and political ramifications of Manus' rise are already being debated. While businesses stand to benefit from increased efficiencies and reduced costs, the specter of job displacement looms large. The fallout from such automation could reshape the labor market significantly—a potential warning for those within the technology sector.
Internationally, Manus' emergence signifies China’s waxing influence over the AI arena, challenging the United States' long-held dominance. With reports indicating the stock market’s immediate response, shares of Alibaba surged 7% following announcements surrounding Manus, reflecting investors' optimism about its potential impact.
Currently, Manus is accessible only to beta testers via invitation, as Monica refines the model's features and adjusts to feedback from the early users. The decision to release it through this restricted access method appears strategic, aimed at ensuring manageable server loads and refining user experience before mass deployment.
Reports from beta users highlight several high-impact use cases already being tested. One user tasked Manus with performing a professional analysis of Tesla, claiming it completed the assignment equivalent to about two weeks of professional work within just one hour. Another user had Manus create a three.js game where players control an airplane without needing guidance, illustrating its potential to handle complex programming tasks efficiently.
While Manus has proven its capabilities on several fronts, industry experts continue to weigh the broader ethical and regulatory issues it raises. The technology behind Manus is based on existing large language models (LLMs), unlike others which are custom-built from the ground up. This raises questions about data privacy and the overarching framework of governance surrounding AI systems.
The growing enthusiasm for Manus is palpable, as reflected in its burgeoning online presence. The chatbot's Discord server has amassed over 160,000 members, and its social media platforms are followed by over 83,000 enthusiasts eager to explore its capabilities.
Nevertheless, the narrative surrounding Manus should also acknowledge a more cautious perspective; early reports had suggested users encountered looping responses or inaccuracies, reminding us of the inherent challenges artificial intelligence continues to face as it evolves.
Nonetheless, the excitement and applications of Manus represent both innovation and risk. The world watches as this latest entrant charts its course through the rapidly shifting waters of AI technology, potentially reshaping industries and livelihoods alike.