The AI world is buzzing about Manus, an “agentic” AI platform that claims to bridge the gap between thought and action. But does it really live up to the hype?
What Is Manus AI?
Manus AI, developed by the Chinese company The Butterfly Effect, promises to be more than just another chatbot. Unlike traditional AI assistants like ChatGPT or Gemini, which rely on constant user input, Manus is designed to make decisions independently.
In theory, this should allow it to perform complex, multi-step tasks – like booking a flight, buying a house, or even creating a video game – without requiring constant human intervention. But does it actually work as advertised?
The Initial Hype: Exclusive Access and Big Promises
Manus AI launched in early March 2024 and quickly gained attention. Within days, its Discord server had over 138,000 members, and invitation codes were selling for thousands of dollars on Chinese resale platforms.
Some early adopters and AI influencers fueled the excitement, praising Manus as a breakthrough. Yichao “Peak” Ji, a research lead for Manus, claimed the platform outperformed OpenAI’s Deep Research model on a key benchmark for AI assistants.
Manus AI also boasts some ambitious use cases, including:
- Real estate scouting – It supposedly considers crime rates, travel times, and other factors when recommending homes.
- Software development – It claims to generate websites from scratch and handle hosting issues.
- Recruitment – The AI allegedly reviews résumés and selects the best candidates based on labor market trends.
However, as more users gained access, it became clear that Manus AI might not be the revolutionary tool many hoped for.
Reality Check: Manus AI’s Performance Issues
Despite the excitement, many beta testers have reported significant problems with Manus AI.
Common Complaints from Users
Users have pointed out several key flaws:
| Issue | Description |
|---|---|
| Frequent errors | The system often crashes or gets stuck in loops. |
| Struggles with basic tasks | Ordering food, booking flights, or making reservations often fails. |
| Factual inaccuracies | Manus sometimes provides incorrect information without citing sources. |
| Broken links | When searching for flights or services, it returns outdated or non-functional links. |
TechCrunch’s Kyle Wiggers shared his own frustrating experience. When he asked Manus to order a fried chicken sandwich, it crashed.
When he requested a business-class flight from NYC to Japan, it returned links to airline websites, some of which were broken. Similar issues arose when he tried to book a restaurant or generate a video game concept.
Why Did Manus AI Go Viral?
Given its flaws, why did Manus AI gain so much attention? A few key factors played a role:
1. Exclusivity Creates Demand
With access limited to a small group of beta testers, invite codes became highly sought after. This exclusivity made Manus feel like a rare, must-have tool.
2. Media and Influencer Hype
Chinese media outlets praised Manus as a major AI breakthrough. Some AI influencers even spread exaggerated claims about its abilities, adding to the frenzy.
3. Comparisons to Established AI Models
Some social media users compared Manus to DeepSeek, a respected AI company in China. However, unlike DeepSeek, The Butterfly Effect hasn’t developed its own AI models—it primarily fine-tunes existing ones like Anthropic’s Claude and Alibaba’s Qwen.
The Bigger Picture: Are We Ready for Fully Autonomous AI?
Manus AI’s struggles highlight a broader question: Are we truly on the brink of fully autonomous AI?
The idea of an AI that can independently make decisions and execute complex tasks is exciting but also raises ethical and regulatory concerns. If an AI makes a costly mistake – say, booking a non-refundable flight to the wrong destination – who is responsible?
What’s Next for Manus AI?
The Butterfly Effect has acknowledged the platform’s flaws and says it’s working to improve Manus by scaling computing power and fixing reported bugs. But whether the AI can deliver on its bold promises remains to be seen.
For now, Manus AI is an interesting experiment in AI autonomy, but it’s far from the game-changer some had hoped for.
Would you try an AI like Manus, or do you think we’re still years away from true autonomous AI?
