🔊 Listen to this ►

A Few Days Of Using Manus AI—What I Discovered.

Manus AI is a new AI agent developed in China, released just a few days ago and it is already described as the world’s powerful fully autonomous AI. This means it can plan, think, and execute tasks independently, setting it apart from traditional AI models that often require human input. Look at AI agents like they are Robots on the Internet, they can operate in this environment, frankly in most cases better than humans. Agent capabilities include data analysis, content creation, customer service automation, and predictive maintenance. We are entering into a new era from chatting to AI to fully self-contained run-it-till-its-done AI agents. See Manus is not a single agent, it is a legion of agents ready to work on the task presented. In most cases, you can walk away and come back with actual work done.

I have had an early viewing and beta invite of Manus because the folks who are building some of this AI over there have successfully discovered the (I thought) anonymous contributions I have made in AI projects and model tunings. They knew I would throw a great deal at this in testing, and I did not hold back. They also know that with mutual respect I will offer to help, something I rarely see with the arrogance of many large US AI companies. I have tested quite a few agent systems, most are not released including OpenAI’s Operator. So my testing is not a one-off for these platforms.

This all sounds great but how good is Manus compared to OpenAI and their forthcoming agents? The research suggests Manus AI performs strongly in the GAIA benchmark, a test for general AI assistants that evaluates real-world problem-solving. Ait scored 86.5% on Level 1, 70.1% on Level 2, and 57.7% on Level 3, compared to OpenAI’s scores of 74.3%, 69.1%, and 47.6% respectively. These numbers indicate it may outperform OpenAI models, though some experts caution that independent verification is limited.

Manus AI’s applications are diverse, covering areas like stock analysis, creating educational materials, comparing insurance policies, travel planning, building software, or even running someone’s online job. For example, it can analyze sales data to provide insights or build websites autonomously. Its ability to integrate with external tools and learn from interactions makes it suitable for automating workflows in businesses.

WARNING, THERE IS PAID CONTENT AT THE BOTTOM. STOP NOW IF THIS IS NOT FOR YOU. YOUR SUPPORT PAYS FOR MY WORK, THANK YOU.

The world of AI is changing by the minute. It is nearly impossible to keep up with the pace. We are here to help with fast and accurate insights. Manus represents the trend of AI originating in China that profoundly disrupts companies like OpenAI as they try to build a long-term case of revenue. This started with the release of DeepSeek R1 and now with the release of Manus. I have done extensive testing on Manus and will give you my insights.

I built a one-prompt game called DogeMan and it wins no awards but is indicative of what can be done with a single prompt. The To-Do list Manus built and used is above. This was not a test of making a game as much as it is a test of the orchestration of this To-Do list used by agents. I am going to update the graphics, sound, and playability soon but here is an early alpha release of DogeMan:https://ouxuuxtk.manus.space

I also point out the Bill Of Materials that Manus uses, Manus is not one AI but a legion of AIs and I found out exactly how and why it works. The open-source versions of Manus are expanding. I have seen 7 pre-released versions, one ironically built by Manus and is quite good.

These insights bring you far ahead of just about anyone working on the most cutting-edge projects in AI. You will be the first to understand this shift and profit from it. If you are a member, thank you. If you are not yet a member, join us by clicking below.

🔐 Start: Exclusive Member-Only Content.

Membership status:

🔐 End: Exclusive Member-Only Content.

~—~

~—~

~—~