Why Manus May Not Signal China’s Next ‘DeepSeek Moment’
Manus, a cutting-edge AI platform that recently launched in preview, is creating a buzz akin to a Taylor Swift concert. Enthusiasts and experts alike are discussing its potential and capabilities, making it a hot topic in the tech community.
Unprecedented Hype Surrounding Manus AI
The excitement around Manus has reached incredible heights. The head of product at Hugging Face described it as “the most impressive AI tool I’ve ever tried,” while AI policy researcher Dean Ball referred to it as “the most sophisticated computer using AI.” The platform’s official Discord server amassed over 138,000 members in a matter of days, with invite codes being sold for thousands on the Chinese resale app Xianyu.
Is the Hype Justified?
Despite the overwhelming enthusiasm, some experts question whether the hype surrounding Manus is warranted. The platform is not entirely original; it reportedly utilizes a blend of existing and fine-tuned AI models, including Anthropic’s Claude and Alibaba’s Qwen. These models enable Manus to perform tasks like drafting research reports and analyzing financial filings.
On its official site, Monica—the Chinese startup behind Manus—showcases ambitious examples of what the platform claims it can achieve, ranging from purchasing real estate to developing video games.
Claims of Superiority Over Other Tools
In a viral video on X, Yichao “Peak” Ji, a research lead for Manus, asserted that the platform surpasses agentic tools like OpenAI’s deep research and Operator. Ji claimed that Manus outperformed deep research on the GAIA benchmark, which tests an AI’s ability to execute tasks by browsing the web and utilizing software.
“[Manus] isn’t just another chatbot or workflow,” Ji emphasized. “It’s a completely autonomous agent that bridges the gap between conception and execution. We see it as the next paradigm of human-machine collaboration.”
Mixed Reviews from Early Users
However, not all initial users share Ji’s enthusiasm. Alexander Doria, co-founder of AI startup Pleias, reported encountering error messages and loops during his tests with Manus. Other users have noted that Manus often struggles with factual questions, inconsistently cites sources, and misses readily available information online.
- Deep Research completed a task in under 15 minutes, while Manus failed after 50 minutes.
- Users have experienced frustrating delays and crashes when attempting simple tasks.
One user recounted their attempt to order a fried chicken sandwich, only to find that Manus crashed after several minutes. On a subsequent try, it located a menu item but couldn’t complete the order.
Limitations in Task Execution
Another user attempted to book a flight from NYC to Japan with clear instructions, but Manus only provided links to airline websites, several of which were broken. When tasked with reserving a table at a nearby restaurant, Manus again failed.
Positive Aspects of Manus
Despite the shortcomings, some users have highlighted Manus’s strengths:
- The platform’s research capabilities and report generation are impressive.
- Its ability to execute background scripts shows significant potential.
- Users appreciate its ambitious approach to AI tasks.
What Lies Ahead for Manus
Monica is aware of the issues users have reported and claims to be working on scaling the platform and addressing these challenges. While Manus currently shows promise, it appears to be a case of hype outpacing actual technological capabilities. For those interested in the latest developments in AI, keeping an eye on Manus’s evolution will be essential.
For further insights into AI advancements and tools, visit our Tech Insights page.
