This New AI Agent Just Changed Everything... (Manus AI Agent)

So there is a new AI agent on the block that is actually shaking up the AI space in a way that I never thought possible. This is the Manus AI agent, which is by a company coming out of China, and they've actually managed to introduce this AI agent, they call it the General AI Agent. And it's able to do a variety of tasks in the browser that are truly, truly effective. Now, the reason that this is, you know, really shaking up the AI space is because this is something that is pretty comparable to OpenAI's deep research and, and their operator. And honestly, in a variety of different tasks, it seemingly performs better. So in this video, I'll show you guys exactly what it's able to do. The drawbacks, limitations, of course, the pros, and the variety of things going on. So one of the first things actually show is they actually show the use case of the agent handling multiple files, dealing with Internet research, dealing with multiple tools, and then coming to a really nice conclusion at the end. So this is called resume screening, and this is what we can see.

Let's start with an easy one. In this example, we'll ask Manus to help screen resumes. I've just sent Manus a zip file containing 10 resume documents. Since each Manus session has its own computer, it can work like a human, first unzipping the file, then browsing through each resume page by page and recording important information to documents. MANIS works asynchronously in the cloud, which means you can close your laptop anytime and MANIS will notify you when everything is complete. Of course, you can also give Manus new instructions at any time here. I've sent Manus five more resumes. After carefully reading all 15 resumes, Manus provides its ranking suggestions along with candidate profiles and evaluation criteria as supporting materials. This is pretty good, but I prefer a spreadsheet. Let's have Manus create one. Manus has its own knowledge and memory, so it can teach Mannis that the next time it handles a similar task, it will deliver a spreadsheet right away. In this example, we'll have Manus conduct some research. It needs to filter New York properties based on multiple criteria. For complex tasks, MANIS first breaks them down and creates a to do list. MANIS begins by searching and carefully reading articles about the safest neighborhoods. Then Manus researches middle schools in New York. Next, Manus writes a Python program to calculate my budget. Based on my budget, Manus filters listings on real estate websites. Finally, combining all the information gathered, Manus writes a detailed report and compiles all the resources.

We have the Manus agent deeply analyzing Tesla stock and one of the cool things about this is that it automatically is able to get financial data. I'm presuming that it is hooked up to a bunch of different APIs, but we can see that this is something that has access to a range of different tools, allowing it to do just a lot more than an actual chatbot. Now of course as always with these replays, they are in two times speed because the Manus AI agent isn't as quick as OpenAI's, but it certainly seems to be a little bit more effective in some areas. So such as financial analysis, where it's able to pull data, it's able to make these charts and it's able to analyze across a variety of different things. And one thing that I do like about this is that we can clearly see exactly the thought process and of course everything that is going on now, of course right here you can see that this is something that is finalized. All of this data is right here. And this is something that is probably going to be truly, truly valuable to the financial industry. Now I not sure what kind of tasks are going to be automated within the finance space, but I do know that there is of course a lot of research, a lot of specific research, things that you know you're going to want to do, especially if you're looking for market updates. And I think things like this where you can easily say, hey, I need you to do xyz, clearly show us the value of agents and you're basically going to be able to do a lot more research easily, especially if you're in the finance space. So this is going to be a valuable, valuable tool. Rather than browsing tons of websites and getting all the data and writing them all down, you can easily have all of this information ready at your fingertips for digestible analysis. There was another one right here where you've got the comparative analysis of travel insurance policies. You can see what the agent is doing. It decides to analyze the travel insurance policies that you provided, highlights the difference, researches them and then overall you can see that this agent is able to do a variety of different things to order, you know, basically just make things even better. And you can see that it had, you know, an entire list there, which is really cool. Then it's going through the list, analyzing everything and ensuring that your outcome is exactly what you need. And I think like I said before, this is going to be something that basically just means that a lot of things that would take hours simply just take minutes now. So this is of course going to speed up a variety of different industries because we now have agents on our behalf that are doing things that are going to be really, really useful. You can see, can see right here, it's able to get all of this information, all of this methodology, provide us with the key insights at a fraction of the time that it would take us. And there are so many tasks like this that most people don't realize do take a ton of time. Another thing that the managed AI agent was able to do was get leads for you. And I think this is going to be really, really valuable, especially for people conducting research in the business space. Getting leads is probably one of the hardest things that many businesses do suffer with. So I think that now that businesses are going to be able to potentially get leads from, from this AI tool and it's able to scrape many different websites, get the emails, get their information and simply be able to contact them. And I do think that OpenAI deep research can do this now, but this is just a demonstration of what MANUS is able to do when it needs to do a certain task. And like I said before, there are a variety of many, many different use cases which I personally find really valuable. And I do think that this is something that OpenAI should have done because, because OpenAI operator, when they showcased the agent, they didn't really showcase too many really good use cases. And I think that this use case here, like the B2B Air Company's dashboard where you can see all of these companies, their locations, you can visit all their websites, you can have all their contact details right here based on some specific details. For example, if we do come over to OpenAI's page where we do see deep research, we do see some use cases here. For example, you've got medical research like here. It talks about how GPT4O is a lot different to deep research, UX, design, shopping and general knowledge. And I do think that there simply aren't enough use cases here. Like they only show 1, 2, 3, 4, 5, 6 use cases, which is of course a decent amount, but it's not that much in terms of actually getting people to creatively think about how you can use this. Whereas you can see with the use case gallery right here, you've got the featured tab, you've got the research tab, you've got the life, you've got data, analysis, education, productivity, wtf. There are just huge, huge, huge amounts of different use cases and, and you can also see what people are trying to do as well on Twitter. If you go on Twitter, you're able to see a variety of different use cases that people have innovatively created. So I think this is going to be something that also allows people to explore things that they may not have thought of before. And I think that's really vital, which is why this company is having this success out of the gate that they are having now in the Gaia benchmark, which is essentially just the general AI agents benchmark for benchmarking where AI agents are, which we can see that they've actually achieved state of the art. And I do think that this is super, super interesting because for the longest time OpenAI has always been state of the art in pretty much every category. But considering the fact that AI agents are somewhat of a new category and companies are managing to innovate on that space, I think considering the fact that they are quite narrow and niched and focused down, it is possible for OpenAI to get beat in certain categories like this Gaia benchmark, but which is one where OpenAI falls short. Now it doesn't fall short by a large margin, which is of course what you'd might expect, but the margin is significant enough to the point where people may reconsider paying for OpenAI's Pro subscriptions. It seems that this year there are many new companies coming out of the woodwork to challenge OpenAI's status as potentially the leading company in the AI agent space. Now this isn't a video dunking on OpenAI at all. I'm actually still pretty bullish on the future technologies, but. But it does go to show what a competitive space can do for the end user. At the end of the day, I'm pretty sure this is going to make sure that OpenAI are on their best behavior in terms of getting their products out quicker, faster and more efficient. Now there was also this that was on Twitter. This is actually not Manus. So if you've seen this clip floating around the Internet where it's apparently the Manus AI agent, this is actually not the Manus AI agent on cell phone devices. So you can see right here, you can co founder of Manus AI here. Thank you for your interest in Manus, but this video is definitely not Manus. Our official account is manusaihq and manus is a web based AI agent like the OpenAI operator deep research and not a VNC for smartphone. So I wanted to include that. Now what about those comparing Manus to Deep Research? Let's actually take a look at what they've said so for Manus versus Deep Research. We can see that this person, this MD Derya Unut Mazz actually says that he's going to share his follow up to the test that he actually conducted comparing the Manus AI to OpenAI's deep research. He talks about how this prompt was highly complex, evaluating an ME CFS disease, generating new hypotheses and identifying neural therapeutics using their recent multi omnics AI paper preprint as the initial context. So he talks about how this condition is a highly debilitating chronic disease similar to Long Covid affecting millions of people. And he studied this condition for over a decade and has made several discoveries and contributions to the field. So he's deeply familiar with it. And he says, let me clearly state from the outset, these two reports generated in a matter of hours could potentially accelerate our understanding and treatment of this disease by months or even years. People will need to recognize just how profoundly AI can impact the advancement of biomedical research. And, and this is where he talks about the fact that, you know, a brief summary and review of the comparisons. Deep Research generated an approximately 12,000 word analysis and report which as expected was outstanding. And surprisingly, manusai produced a report that was shockingly good as well, almost as twice as long, 21,000 words, and even slightly more innovative and comprehensive in hypotheses, generation and therapeutic suggestions than Deep research. However, Manus AI took four times longer to complete the analysis, which was approximately one hour versus the DEEP researchers 15 minutes, and its report could have been somewhat shorter and tighter. Both models had significant overlap in their analyses, but each also uniquely highlighted important references not mentioned by other, suggesting a complementary potential that I'll explore next. And in summary, he's basically saying that Manus AI is pretty shockingly comparable to to OpenAI's deep research. And apparently people are saying that this is another deep seek AI moment. And I gotta be honest with you guys, it certainly is, because when we actually take a look at some of the other use cases on the webpage, they are truly, truly impressive. So what do you guys think about the Manus AI agent? Do you guys think this is super effective or do you guys think that this is not that effective? I think it is, of course another wake up call to the AI industry that China is ramping up their AI efforts and seemingly they might even surpass the Western counterpart.