China’s AI industry is again making headlines with the launch of Manus AI, a fully autonomous AI agent developed by the startup Monica. This new AI model claims to handle complex tasks with little human input, sparking excitement and skepticism. While some call it China’s next DeepSeek moment, others believe it’s overhyped. Here’s everything you need to know about Manus AI and why it’s causing a stir in the AI world.

What is Manus AI?

Manus AI is designed to perform entire workflows autonomously. Unlike traditional AI chatbots that require step-by-step instructions, Manus AI can:

• Sort and rank job applicants.
• Analyze stock market trends.
• Scrape web data for insights.
• Build fully functional websites from scratch.

The AI runs in the cloud, meaning it continues working even when users log out. According to developers, Manus outperforms OpenAI’s Deep Research model on the GAIA benchmark, a widely used AI evaluation tool. However, access to Manus is invite-only, adding to the mystery and hype around the product.

Is Manus Truly Revolutionary?

While its creators claim that Manus AI is a breakthrough, some AI researchers believe it’s not entirely built from scratch. Early users noticed that Manus’ responses contained patterns specific to Claude 3.5 Sonnet, an AI model developed by Anthropic.

Manus’ co-founder, Yichao’ Peak’ Ji, later confirmed that the AI currently operates using a mix of Claude 3.5 Sonnet v1 and fine-tuned versions of Alibaba’s Qwen models. This suggests that Manus isn’t an entirely new AI but a well-optimized combination of existing technologies.

Manus vs. DeepSeek: A Fair Comparison?

With its bold claims, many have compared Manus to DeepSeek, the Chinese AI model that shook the industry earlier this year. DeepSeek’s launch proved that China could develop AI models rivaling Western giants despite U.S. sanctions on advanced chips. It even led to a $600 billion sell-off in Nvidia stock.

However, some experts argue that Manus is taking AI even further.

AI policy researcher Dean Ball commented on X:

“DeepSeek was about replicating American AI capabilities. Manus is actually advancing the frontier.”

Unlike DeepSeek, which focuses on improving general AI capabilities, Manus is designed to be a practical AI agent capable of automating real-world workflows.

Concerns Over Manus AI

Despite its impressive demo, many early testers have reported issues, including:

• Execution failures: The AI sometimes gets stuck in loops, unable to complete tasks.
• Factual errors: Manus has been found to make mistakes in data analysis.
• Plagiarism concerns: Some suspect it scrapes content from existing websites without proper attribution.

Another primary concern is data privacy. AI researcher Luiza Jarovsky has questioned where Manus stores its data and whether Chinese authorities can access it. Since Monica, the company behind Manus has links to Butterfly Effect, a Singapore-registered entity, some believe there could be hidden risks. Jarovsky asked:

“Where are Manus’ servers located? Does user data get transferred to China?”

These concerns highlight the growing debate over AI security and governance.

Is Manus AI a Game-Changer?

As the AI world watches closely the question remains: Is Manus AI a groundbreaking advancement or another overhyped tool? While it shows great potential, its true capabilities will only become apparent as more users gain access and test its real-world applications.

Final Thoughts

Manus AI has undoubtedly captured the industry’s attention, but its actual impact remains to be seen. While some experts believe it marks a new era in AI automation, others argue it’s a refined version of existing technology. What’s clear is that China’s AI sector is rapidly evolving, and models like Manus and DeepSeek are proving that the global AI race is intensifying.

FAQs About Manus AI

1. What makes Manus AI different from other AI models?
Manus AI is designed to work autonomously, handling complex tasks without constant human input, unlike traditional chatbots that require step-by-step guidance.

2. Is Manus AI truly independent or based on existing models?
Manus AI uses a combination of Claude 3.5 Sonnet and fine-tuned Qwen models rather than being a completely new AI built from scratch.

3. How can I access Manus AI?
Currently, access to Manus AI is invite-only, making it exclusive to select users for testing.

4. What are the biggest concerns with Manus AI?
Some concerns include execution errors, potential plagiarism, and data privacy issues, particularly regarding whether user data is stored in China.

5. Will Manus AI change how we use artificial intelligence?
While it has the potential to redefine AI automation, its long-term impact depends on further development, reliability, and adoption by industries worldwide.

What do you think? Is Manus AI a true breakthrough or just another hyped-up tool? Let us know in the comments!