Wenchang, China-29 May: a long March-7 Y3 Carrier rocket holding tianzhou-2 cargo ship … [+]
Another “sputnik moment”? Another “deep progress”? Hard. The growth of the Chinese – Deepseek, Groku, now Manus – has created a lot of attention in the media. But Manus he, the last one from China, is far from the novel. Unlike Deepseek, who introduced significant advances, Manus offers nothing revolutionary. It claims autonomy, but in reality, it is merely another great model of language (LLM) that executes the workflow written. The only innovation? It’s from China.
Manus gives average, not intelligence
Like many other remedies, the manus relies on LLM that generate plans based on statistical averages. If I ask chatgpt, Gemini, or depth for help with an article about manus, they will generate an approximated sketch with the predominant market narrative – pronounced themes such as “Breakthrough”, “Hand Universal” and “winning China”. Why? Because that’s what the Internet says. LLM generate average statistical, not knowledge.
Yes, Manus (like any other means) can create a travel itinerary, but will it be quite well? In my ECORNELL certificate program, I introduce the concept of the product of minimal quality. Does the manus meet the sign? It depends – yes, if you treat this summary travel guide like any other. But for fully autonomous booking of travel? I don’t think we’re there yet.
Agents of he cannot make decisions
Decision-making requires the advantage and agency-something he lacks. Get Evans Benedict Analysis for OpenAi’s Deepresearch model for phone data analysis. He produced reliable but essentially flawed knowledge. Why? Because it could not discern what data was most important. Manus will face the same restriction.
Limited workflows – the main success for R2Decide and others
It succeeds when it works within the course of structured, well -defined work. Companies that focus on specific domain-not general applications, all agents-will create real value. Reducing complexity, essentially close cases of use, allows it to function effectively. For example:
R2DecideA company that I founded focuses on optimizing e-commerce conversion. It uses product data, the content created by the user and insights user travel to provide recommendations directed from the one that increases the conversion rates. It works within a narrow, narrow stretch – integrated into the existing work flow. It is not general, the opposite is specialized. However, in A/B tests, it constantly increases revenue for our customers by 5% or more. The broad, unclear agents of it fail – at least for now. He runs well, driven by the workflow, blooms.
Human supervision is missing with manus
The future of it will be – wait for it – people. Human supervision is critical. Most companies struggle with integration and it effectively. Here is a video where I discuss human interaction and health care. While it can improve escalation, human involvement remains essential. As Benedict Evans showed. Openai can research the data and draft a report, but this will be wrong. The future will be companies that open the surface questions where he lacks trust. However this is not what manus is doing.
My Ecornell Certificate “Designing and building it solutions” is an open code course for everyone. To support every student-in addition to the coding experience-I had to build a co-standard of he. Students equipped with an agent of him, then explore and develop products. What students show after course, he is excellent in summary. Prioritize.
Product design it should emphasize two elements: texture and control. He must be directed. The true success of it is not about replacing people; It is about adding them.
Integration: True Challenge
For it to be truly transformer, it must be integrated without effort with external applications, databases and services. Demot of using many manus apps on social media tools are a step in the right direction. Today’s automation platforms – like IFTTT and Zapier – currently offer basic workflow integration, but they will follow suite.
Get Amazon, they recently announced that Alexa will soon be ready for conversations, getting deep into existing ecosystems. Where the market will move. Can Manus reach that level of integration at any time soon? I have my doubts.
First enterprise, customer later
Soon soon he’s autonomous assistants for consumers at all times. Enterprises will approve agents first because they can control setting, manage risks and secure ROI. Main counseling firms such as Accenture and BCG are already removing the solutions of the tasks. Customers will follow – the GPT does not take the world from the storm overnight, and the assistants and it still need many repetitions. If we are there Google and Openai are likely to prevail this space – or a completely new user flow – but not manus with the same old ones.
The success of it is based on application and not “made in China”
Manus’s biggest strength he is not her technology – it’s her brand “Made in China”. The future of it has nothing to do with the bright demonstrations; It is about sustainable, safe and well -integrated solutions. Companies that refine the workflow, integrate effectively and keep people into the loop will determine the industry. Those who follow the light marketing will be left behind.