Beijing, March 15, 2025 – In a bold leap toward the future of artificial intelligence, Chinese startup Butterfly Effect has introduced Manus, an AI “agent” that operates with remarkable independence, making decisions without the need for constant human guidance.
Hailed as the world’s first general AI agent, Manus offers a glimpse into the potential of artificial general intelligence (AGI). On this frontier, machines could rival human adaptability across various tasks.
A team of visionary scientists develops Manus, which stands out from familiar AI tools like ChatGPT or DeepSeek.
While chatbots thrive on detailed prompts and frequent input, Manus leverages a cutting-edge multi-agent architecture—integrating multiple large language models (LLMs) and autonomous software—to tackle complex projects with minimal oversight.
Early adopters have already used it to create everything from interactive video games to fully functional websites, igniting a wave of curiosity and debate online.
A Game-Changer in the Making
Though not publicly available, Butterfly Effect has rolled out a limited batch of invite codes, allowing a select group to test Manus’s capabilities.
The results? A mix of awe and growing pains. Users report that Manus can craft a detailed vacation plan, analyze stock market trends, or even screen job applications—all from a single instruction.
Videos circulating on X showcase its ability to code a retro-style game in minutes, while others highlight its knack for designing sleek websites.
However, it’s not all smooth sailing. Testers have flagged issues like system crashes, infinite loops, and a failure rate higher than ChatGPT’s.
Chief scientist Peak Ji took to X to address these hiccups, calling them “part of the journey” for a tool breaking new ground. “We’re not just building another chatbot,” Ji posted. “We’re redefining what AI can achieve.”
Tech reporter Caiwei Chen, who explored Manus for MIT Technology Review, described it as “a brilliant but quirky collaborator—like an intern with endless energy but occasional blind spots.”
In her hands-on review, Chen noted that while Manus excels at efficiency, it sometimes misinterprets tasks or skips critical steps. “It’s a diamond in the rough,” she wrote, “with untapped potential waiting to shine.”
What Makes Manus Unique?
Unlike traditional AI, which is reliant on a single LLM, Manus operates as a team of digital agents, each handling specific aspects of a task—research, coding, analysis—while communicating seamlessly.
According to Forbes, this allows Manus to adapt its approach on the fly, digging deeper into problems than chatbots typically do.
Data from the GAIA benchmark backs this up, showing Manus often outshines ChatGPT in response depth, though it sacrifices speed for thoroughness, as noted by Tom’s Guide.
Manus keeps working entirely in the cloud even when users log off—a feature that could transform workflows in industries like software development, finance, and beyond.
Imagine assigning it to draft a business proposal or scour the web for market insights, then returning to polished results hours later.
China’s AI Revolution Gains Momentum
Manus is the latest in a string of Chinese AI breakthroughs shaking up the global tech scene. Earlier this year, DeepSeek stunned the industry with a budget-friendly chatbot rivaling Western giants.
Now, Manus could signal the rise of AI agents—tools that act less like assistants and more like proactive partners. This surge underscores China’s ambition to lead the AI race, a topic explored in depth by Reuters.
Yet, with great power comes great responsibility. Experts are sounding the alarm on ethical challenges posed by autonomous AI.
Unlike humans, agents like Manus can work 24/7 without tiring (assuming no crashes), raising questions about job security, data ethics, and accountability.
A recent BBC report on AI advancements warned that such tools could amplify societal risks without robust oversight.
Practical Uses and Future Promise
Butterfly Effect has showcased Manus’s real-world applications on its website, from curating travel itineraries with local recommendations to screening resumes based on custom criteria.
Unlike ChatGPT, which might list options and stop there, Manus takes initiative by booking mock flights or ranking candidates by skill match.
Online, reactions range from amazement to cautious optimism. One X user shared a demo video of a puzzle game Manus built from scratch, while another griped about a stalled task: “It’s brilliant until it isn’t.” These mixed reviews highlight the stakes: Manus could redefine AI—or stumble as an ambitious experiment.
What’s Next?
As Butterfly Effect fine-tunes Manus, the tech world is buzzing with speculation. Could this be the stepping stone to AGI, or just a flashy proof-of-concept? For now, it’s a bold experiment pushing AI beyond its comfort zone. Stay tuned for updates—and maybe an invite code—as this story unfolds.
