One evening in Shenzhen, a group of software engineers huddled in a dimly lit workspace, typing furiously as they monitored the performance of a new AI system.

The hum of servers mingled with the glow of high-resolution screens. They were testing Manus, an AI Agent capable of independent thought and action.

In just a few hours, its arrival on March 6 will send shockwaves through the global AI community, reigniting a heated debate that has simmered for decades: What happens when AI stops asking for permission and starts making its own decisions?

Manus is more than just a chatbot. It is the world's first autonomous AI Agent, a system that not only assists humans but also replaces them.

From analyzing financial transactions to scanning candidate resumes, Manus navigates the digital world without human supervision, making decisions so quickly and accurately that even seasoned professionals can't keep up.

In essence, it is a digital scientist trained to perform multiple tasks in different industries without hesitation.

manus founder
Yichao "Peak" Ji, one of the founders of Manus AI. Screenshot.

But how did China, often seen as lagging behind the US in fundamental AI research, come up with something that Silicon Valley still considers theoretical? And more importantly, what does it mean for the balance of power in AI?

“DeepSeek Moment”

In late 2024, DeepSeek released a low-cost, high-performance AI model that rivaled OpenAI’s GPT-4. It was the AI ​​world’s “Sputnik moment,” and the first tangible sign that Chinese researchers were closing the gap in the capabilities of large language models (LLMs).

However, Manus represents something completely different: it is an Agent – ​​an AI system that can think, plan, and act independently.

That’s what sets Manus apart from its Western competitors. While ChatGPT and Google Gemini require human input of prompts, Manus doesn’t need guidance. Instead, it’s designed to work on its own, learning new information and adapting its approach proactively.

For example, when presented with a set of candidate profiles, Manus not only ranks them but also reads each profile one by one, extracts relevant skills, compares them with market trends, and makes optimized hiring decisions using a self-generated Excel file.

When asked to “find an apartment in San Francisco,” Manus not only looks at search results, but also considers crime rates, rental trends, and even weather patterns to come up with a shortlist that’s right for the user.

Invisible Staff

Manus can be thought of as an invisible assistant that uses the computer just like you do, except for one thing: it never gets tired. The key is its multi-agent architecture.

Instead of relying on a single neural network, Manus acts as a director overseeing a team of specialized sub-agents. When given a complex task, it breaks the problem down into its components, assigns each agent to it, and checks on progress.

This architecture addresses multi-step workflows that previously required multiple AI tools working together.

Another difference lies in its asynchronous, cloud-based operation. Traditional AI assistants require active interaction from the user, whereas Manus works in the background and only notifies the user when the results are ready.

Manus signals a shift from AI as an assistant to AI as an independent agent. Rowan Cheung, a tech writer, tested Manus and asked it to write a biography about him and build a personal website.

In just a few minutes, the AI ​​scans social media, extracts professional highlights, creates a neat bio, writes a website, and publishes it. It even handles hosting without any additional input.

For AI developers, it is the Holy Grail, a system that not only generates information but also applies it, fixes errors, and refines the results. For workers, Manus is an existential crisis.

Shock to Silicon Valley

For years, AI revolved around the big American tech companies: OpenAI, Google, Meta. It was assumed that whoever created the most sophisticated chatbot would control the future of AI. Manus shattered that assumption.

Manus is a completely new category of AI that shifts the focus from passive assistance to autonomous action. What's more, it's a product of China.

So it’s making Silicon Valley uneasy. China’s aggressive push into autonomous systems will give it a first-mover advantage in key areas.

They fear that Manus represents the industrialization of AI: systems so effective that businesses will soon be forced to replace humans with AI.

But Manus also raises questions about governance and ethics. What if an AI Agent makes a financial decision that costs a company millions of dollars? Or if it executes an incorrect command, will it have real-world consequences? Who is responsible when an unsupervised, automated system makes a mistake?

So far, the biggest question is not whether Manus will “live up to its name,” but how long it will take for the rest of the world to catch up.

The era of autonomous AI Agents has begun, and China is leading the way. In the meantime, we may have to rethink how we work, innovate, and compete in a world where AI is no longer an assistant.

(According to Forbes)

Another Chinese AI that is as 'hot' as DeepSeek Manus, an AI Agent that was just launched in China, has attracted attention thanks to its ability to handle complex tasks. Many people expect Manus to achieve the same success as DeepSeek.