Manus ignites AI craze! A-shares hit the daily limit up again, the next DeepSeek?

The field of artificial intelligence is once again experiencing a 'thunderbolt'.


At that time, an AI product called Manus suddenly emerged. According to its official introduction, this is the "first universal AI agent" that can independently think, plan, and execute various complex tasks, and directly deliver complete task results.


It looks like it can really help people do things, which attracts a group of people at home and abroad who want to try it out as soon as possible. However, while Manus was still undergoing internal testing, most of them were unable to avoid the "bad luck" of staying up all night waiting for the invitation code to fail.


People who woke up and saw Manus flooding the screen were shocked by its invitation code being sold for hundreds or even tens of thousands of yuan on second-hand trading platforms, and anxiously lined up in the waiting queue.


In addition, this is also a "phenomenal work" created by a Chinese team, so it is no wonder that netizens called this night "another sleepless night in the technology industry after DeepSeek".


The A-share market also experienced another surge of limit up. As of the close on March 6th, over 20 AI intelligent agent concept stocks have risen by the limit up or more than 10%.


The next DeepSeek?


The selling point of "universal AI agent" is the core that Manus can attract widespread attention this time.


What is an AI agent? Simply put, compared to conversational AI products that provide responses, suggestions, or answers, AI agents go further in their ability to make autonomous decisions, execute tasks, and even complete operations.


This involves longer and more complex thought chains, as well as the ability to autonomously call a range of tools.


For example, in the official video where Manus demonstrated a case of helping people choose a house, it can break down the complex task of choosing a house into multiple to-do items based on the user's buying needs, such as location in New York, safe environment, high-quality educational resources, budget, etc., and collect information and generate corresponding analysis for each to-do item; Even they will use programs to crawl the price information of real estate websites for users, in order to screen out alternative options for them. Ultimately, Manus will provide users with a detailed report on community safety, budget, and recommended property listings.


Image/Manus official introduction video


For example, if Manus is asked to help screen and rank candidate resumes, it can automatically decompress compressed files uploaded by users, mark key points after carefully reading all information, and finally provide users with not only ranking suggestions, but also grading based on candidate experience and other factors, which can be presented in a visual way.


This makes Manus look like a worker who can help the boss think more.


The reason why it is called a universal AI agent may be because the development team is confident in Manus' ability to adapt to different fields and scenarios.


From all the cases displayed on the Manus official website, it can not only help develop travel plans and compare insurance policies for different professions such as teachers and e-commerce professionals, but also handle a series of relatively more professional tasks such as course design and online store management analysis.


Meanwhile, according to Manus' official website, its score in the GAIA (General AI Assistant Assessment) benchmark test exceeded OpenAI's Deep Research feature, achieving SOTA, which means that the performance of the model or algorithm has reached the current best level.


Image/Manus official website


In addition to its functionality and performance, another noteworthy aspect of Manus is that, like the previous viral DeepSeek, it is a product developed by a Chinese team.


Public information shows that Xiao Hong, the founder of Monica.im, the team behind Manus, graduated from Huazhong University of Science and Technology with a major in software engineering and is a serial entrepreneur. Xiao Hong has successively launched official account editor companion assistant for B-end users and private domain traffic management platform companion assistant, as well as monica, a browser plug-in that provides C-end users with chat, translation and other functions by integrating multiple large models.


However, it is worth mentioning that while it is difficult to obtain a single size for Manus, doubts have also begun to emerge. Some AI industry practitioners believe that Manus' invitation code is more like a hunger marketing tactic, and the overwhelming "feature hype" is more like official hype.


In response to this, Manus partner Zhang Tao stated on social media that the team has never opened any channels for obtaining invitation codes through payment, and has never invested any marketing budget; Due to limited system capacity during the internal testing period, priority will be given to ensuring the core experience of existing users and gradually releasing invitations in an orderly manner.


At the same time, the aforementioned AI industry practitioners also told China Newsweek that even if it is hype, it does not mean that Manus does not have real strength. But everything still needs to wait for it to open up for testing, to see if it can withstand more tests and whether it really leads as it claims


Expectations and Risks


Although the vast majority of people have not yet been able to use Manus, it has not stopped the flames of the stars from spreading towards the A-share market.


Starting from the morning session of March 6th, the concept of AI intelligent agents exploded across the board. Among them, multiple stocks hit the limit up, with KuTe Intelligent opening with a "20cm" limit up, and Zhejiang Wenxin Internet and Nanxing Stock opening with a "10cm" limit up. By the midday closing, the daily limit up trend had intensified. Cube Holdings, Hande Information, Zhongyi Technology, Zhiyuan Internet, and others all recorded strong limit up, while Focus Technology, UFIDA Network, and others also hit the limit up.


As of the close of the day, there were over 20 stocks with a daily limit up or an increase of more than 10%.


In the view of Liu Chunsheng, an associate professor at Central University of Finance and Economics, the main reason for the explosion of AI intelligent agent concept stocks is that the release of universal AI agents marks the upgrade of AI from "providing answers" to "autonomously executing tasks", which is both a technological breakthrough and an industry trend. Institutions predict that 2025 will be the first year of commercialization for AI agents, and the market size is expected to increase significantly


According to Deloitte's forecast, by 2025, 25% of enterprises using generative AI will deploy AI agents, and this proportion will increase to 50% by 2027. Capgemini, a management consulting firm, also pointed out in a report that the majority of enterprises (82%) plan to integrate AI agents by 2026.


As for the Chinese market, according to IDC's forecast, by 2026, 50% of China's top 500 data teams will use AI agents to achieve data preparation and analysis, and these data teams will become important organizers and coordinators of related work.


Nowadays, the emergence of Manus has given the market a basis for the implementation of AI Agent applications, and market confidence has increased accordingly.


In fact, AI agents have indeed become the future that many industry players are exploring.


OpenAI CEO Sam Altman once said, 'Agents will be the next major breakthrough.'; According to foreign media reports, it is planning to launch three "doctoral level" AI agents to handle complex tasks in academic research and software development, with monthly fees potentially ranging from $2000 to $20000. Robin Lee, the founder, chairman and CEO of Baidu, also publicly stated that "Agent is our most promising AI application development direction".


But Liu Chunsheng reminds investors to pay attention to the investment risks involved.


On the one hand, AI agent technology is still in its early stages, and there is still uncertainty in its implementation, with risks such as research and development failures and difficulties in adapting to different scenarios; On the other hand, some companies only announce their integration or cooperation, with limited actual contributions, and may face a pullback in the future Liu Chunsheng said, "At the same time, short-term stock price speculation may overdraw future performance, and the industry competition is fierce, making companies lacking core technology easy to be eliminated


Qu Fang, an investment advisor at Wanlian Securities, also stated that this concept is still in its early stages and needs to be validated for future market-oriented applications and practical effects. Investors also need to objectively and rationally analyze it, especially paying attention to changes in the market penetration rate of its future products.


Source: China Newsweek