Tech Desk 
Mar 06, 2025
In a remarkable feat representing China's unwavering potential to stand out in the innovative realm of artificial intelligence (AI), a Chinese firm has disclosed the world's first fully autonomous AI agent, Manus.
Monica.im, the developer behind Manus, the self-controlled AI agent, is designed to handle complex tasks, which most other conventional AI chatbots struggle to oversee, as reported by The Express Tribune.
What sets Manus apart from typical AI tools is that instead of giving suggestions or directly answering user queries, it commits self-regulating, comprehensive task completion.
The mechanism behind the AI agent is driven by a multi-signature (multisig) approach equipped with multiple independent models. The biggest development around Manus is that developers are planning to open-source parts of it, especially the constituent facilitating in inference, later this year.
The company shared a four-minute video teasing Manus autonomously executing tasks from planning to completion. Part of the demonstration involves the AI agent assessing aspiring candidates for a reinforcement learning algorithm engineer position by sifting through resumes to find essential requirements for the role.
Manus is said to have raised the bar of state-of-the-art (SOTA) performance benchmarks across all difficulty levels in the GAIA assessment, a test designed to determine AI assistants' capabilities.
