When Large Language Models first emerged, they quickly became the talk of the town. Over the years, we’ve witnessed remarkable advancements, with Google’s Gemini 2.5 Pro surpassing all previous expectations in nearly every category in which LLMs are tested. While there is still plenty of room for improvement—and improvement is definitely on the horizon—the pace of progress has slowed a bit.
From the average user's perspective, who mainly uses LLMs as chatbots, the changes are less striking. However, the story is different for developers who use LLMs to assist with code writing. For them, the improvements from one model to the next are still highly noticeable. But that’s not the focus of this discussion.
With the advent of LLMs, people realized they could build systems that use an LLM as a "brain". This led to the creation of what we now call AI agents. In the previous article in this series, we explained what AI agents are and how they work at a high level. In this article, we will explore some of the most popular AI agents available. We will also demonstrate how we can use an AI agent to solve a task from start to finish, simply by describing what we want it to do in just a few sentences.
How to Explore Existing AI Agents on the Web
AI agents have become so popular lately that it feels like a new one pops up every time you scroll through a tech article or news feed. Of course, this perspective is a bit skewed. People actively searching for these tools are often professionals deeply immersed in the field, much more so than the average person. Still, it’s fair to say that AI agents are the hot topic right now. Everyone seems to be jumping on the bandwagon, trying to claim their share of the action.
In our previous article, we highlighted some of the most popular AI agents currently available to the public. We briefly mentioned agents such as AutoGPT, MetaGPT, and others. While we discussed how these AI agents work from a theoretical perspective, we didn't demonstrate how to actually use them. In this article, we will focus on providing a hands-on demonstration of how to use two highly popular online AI agents:
- AgentGPT
- Manus AI
Both of these are online AI agents that offer you the ability to test them for free. Therefore, you should be able to follow along with this article without running into any issues.
Article continues below
Want to learn more? Check out some of our courses:
What Is AgentGPT
AgentGPT is an AI agent developed by the Reworkd team, launched in early 2023. It quickly gained attention for empowering users to create, customize, and deploy AI agents capable of independently planning, reasoning, and executing complex tasks. Unlike traditional AI assistants, AgentGPT does not require constant human input. It excels at managing tasks on its own. The process is simple: it breaks down any problem, regardless of size or complexity, into manageable subtasks. This approach allows the agent to handle intricate projects effectively, without the need for continuous oversight.
The open-source nature of AgentGPT has fostered a vibrant community, as seen in its rapid rise to over 25,000 GitHub stars within just a few months. Notably, there is also a consumer-facing application designed for non-technical users. In this article, we will be using that version to solve some fairly complex tasks. In upcoming articles in this series, however, we will move beyond these online platforms and focus on implementing systems on our own.
The AgentGPT online platform is designed to simplify the process of creating, configuring, and deploying autonomous AI agents. The platform features a clean, user-friendly web-based dashboard that allows users to navigate and interact with the system effortlessly.
As can be seen, users can easily create agents by assigning a name, defining a role, and setting a goal or objective. The system then leverages advanced language models to autonomously plan, reason, and execute tasks. In this article, we will be testing the free tier, which uses GPT 3.5 Turbo, while the paid tier uses GPT 4.
While these models perform well overall, they are not State-Of-The-Art models when it comes to certain tasks, such as programming. Therefore, we will limit ourselves to working on tasks that do not require the agent to generate highly advanced code for us. For beginners, the easiest way to create an agent is to choose one of the available templates. They can be accessed by clicking on the Templates button in the sidebar.
Templates are a new feature, so at the moment, there is a total of 20 templates available. Each template represents customizable, ready-to-deploy agents, organized into several categories for easy navigation.
Let's go ahead and use one of these templates. To be more precise, let's use the IndustryGPT template. When you click on a template, you'll be redirected to the home screen. However, you'll notice that both the Name and Goal sections will now be filled with text. The name section will be the name of the template, and the Goal section will contain a default goal, which is:
Conduct an in-depth examination of the ClimateTech industry, detailing its current market status, emerging trends, significant challenges, and opportunities. Make sure to include data and statistics, a list of major players, a forecast for the industry, and how current events or developments could influence it.
This goal serves as a guide for how you should structure the task you assign to your model. For our purposes, let's keep everything identical, but analyze the EdTech industry instead. In addition to defining the task, we can also specify which tools the agent should use to solve it. Clicking on the Tools button brings up a menu where users can pick which tools will be used for execution:
By default, the agent can generate images using AI, search for information online, and write and review code. Since we have no relevant data to provide, we will let the agent run using these default settings. While the agent is running, you can easily track its train of thought and monitor which tasks it is performing in real-time:
After running, the agent returns the output in text format. Here's the breakdown it generated based on the prompt we provided:
The EdTech industry is currently experiencing significant growth and evolution, driven by various segments within the market. In 2022, the hardware segment dominated the market with over 40.90% of global revenue share, indicating the importance of physical devices in educational technology [1]. However, the software segment is expected to witness a Compound Annual Growth Rate (CAGR) of 14.5% from 2021 to 2028, signifying a shift towards digital solutions in education. The content segment is also projected to grow at a CAGR of 16.8% during the same period, highlighting the increasing demand for educational content delivery platforms.
The global EdTech market size was estimated to be USD 142.37 billion in 2023, with a forecasted growth rate of 13.4% from 2024 to 2030 [2]. By 2024, the market had expanded further, reaching a value of USD 169.37 billion and growing at a CAGR of 17.10% [3]. This continuous growth trajectory showcases the increasing adoption of technology in educational settings, both in traditional institutions and online platforms.
The K-12 EdTech segment emerged as the leader in 2023, commanding a significant revenue share of over 39% [4]. The market is further categorized by sector into Preschool, K-12, Higher Education, Corporate Training, and Others, with the K-12 segment holding a dominant position [5]. This segmentation reflects the diverse needs of different educational levels and the targeted solutions provided by EdTech companies.
Looking ahead, the EdTech market is poised for substantial growth. It is projected to reach USD 250.2 billion in 2024 and is expected to grow at a CAGR of 11.86% to reach USD 721.2 billion by 2033 [6]. This forecast indicates a lucrative opportunity for businesses to research and analyze emerging trends in the EdTech industry, particularly focusing on personalized learning, virtual reality, and artificial intelligence.
Personalized learning is a key trend in EdTech, aiming to tailor educational experiences to individual student needs and preferences. This approach involves the use of adaptive learning platforms, data analytics, and AI algorithms to deliver personalized content and feedback to students. Companies like Knewton and DreamBox Learning are leading the way in personalized learning solutions.
Virtual reality (VR) is another emerging trend in EdTech, offering immersive and interactive learning experiences. VR technology allows students to explore virtual environments, conduct simulations, and engage in hands-on learning activities. Companies like zSpace and Nearpod are pioneering VR applications in education.
Artificial intelligence (AI) is transforming the EdTech landscape by enabling personalized learning experiences, automating administrative tasks, and providing intelligent tutoring systems. AI-powered tools like Squirrel AI Learning and Carnegie Learning are revolutionizing how students learn and teachers instruct.
In conclusion, personalized learning, virtual reality, and artificial intelligence are shaping the future of education technology. These trends are enhancing student engagement, improving learning outcomes, and revolutionizing traditional teaching methods. As the EdTech industry continues to evolve, it is crucial for educators and stakeholders to embrace these technologies to unlock the full potential of digital learning.
The EdTech industry is currently experiencing rapid growth and innovation, with numerous opportunities for further expansion. However, there are significant challenges that the industry must address to maintain its success. One major challenge is data privacy concerns. As EdTech companies collect and store large amounts of sensitive data from students and educators, ensuring the security and privacy of this information is crucial. Implementing robust data protection measures and complying with regulations such as GDPR are essential steps to address this challenge.
Another challenge is the integration of EdTech platforms with traditional education systems. Many schools and educational institutions have been slow to adopt new technologies, leading to issues with compatibility and usability. Collaboration between EdTech companies and educators is key to overcoming this challenge, as it allows for the development of solutions that meet the needs of both parties.
Scalability is also a significant challenge for the EdTech industry. As the demand for online learning continues to grow, EdTech companies must effectively scale their operations to meet this demand. Investing in infrastructure, technology, and human resources is essential to ensure that EdTech platforms can handle increased user traffic and deliver a seamless learning experience.
Overall, addressing these challenges will be crucial for the continued success and growth of the EdTech industry. By prioritizing data privacy, fostering collaboration with traditional education systems, and focusing on scalability, EdTech companies can overcome obstacles and seize opportunities for innovation and expansion.
To explore opportunities in the EdTech sector, various avenues for growth and expansion can be considered:
- Expanding into new markets: One opportunity in the EdTech sector is to expand into new markets, particularly in developing countries where access to quality education is limited. By customizing products and services to fit the specific needs of these markets, EdTech companies can tap into a largely untapped customer base. For example, offering low-cost or free educational resources in local languages could appeal to a wider audience.
- Partnerships with schools and universities: Collaborating with schools and universities can provide EdTech companies with valuable insights into the educational system and help tailor their products to better align with curriculum requirements. Partnering with educational institutions can also help gain credibility and trust among educators, students, and parents. By offering exclusive discounts or access to premium features for institutions, EdTech companies can establish long-term relationships and secure a steady revenue stream.
- Innovative product development: Investing in innovative product development is crucial for staying competitive in the rapidly evolving EdTech industry. This could involve implementing cutting-edge technologies such as virtual reality, artificial intelligence, or gamification to enhance the learning experience. Developing personalized learning solutions that cater to individual student needs can also set companies apart from their competitors. By continuously improving and updating their offerings, EdTech companies can attract more users and retain existing ones.
Overall, exploring opportunities in the EdTech sector requires a combination of market research, strategic partnerships, and innovative thinking to capitalize on the industry's growth potential. By identifying and leveraging these opportunities, EdTech companies can not only expand their reach but also make a meaningful impact on the future of education.
To conduct an in-depth examination of the EdTech industry, we first need to understand its current market status. As of now, the EdTech industry is experiencing significant growth due to the shift towards online learning caused by the COVID-19 pandemic. According to recent data, the global EdTech market size is projected to reach $404 billion by 2025, with a Compound Annual Growth Rate (CAGR) of 16.1%.
Some emerging trends in the EdTech industry include the rising popularity of personalized learning solutions, the integration of artificial intelligence and machine learning technologies, and the increasing focus on skills-based education rather than traditional degrees.
However, the industry also faces significant challenges such as data privacy concerns, the digital divide in access to technology, and the need for effective teacher training in utilizing EdTech tools.
In terms of opportunities, the EdTech industry has the potential to expand further into emerging markets, collaborate with traditional educational institutions for blended learning approaches, and innovate through gamification and virtual reality technologies.
Now, to conduct a comparative analysis of major players in the EdTech industry – Coursera, Udemy, and Khan Academy, we will look at their market share, revenue growth, and key strategies for market expansion.
Coursera, a leading online learning platform, holds a significant market share due to its partnerships with universities and corporations worldwide. Its revenue has been steadily growing, with a focus on offering accredited courses and degrees to learners. Coursera's key strategy for market expansion is to continue forming strategic partnerships and expanding its course offerings in emerging fields.
Udemy, another major player in the EdTech industry, focuses on offering a wide range of courses taught by industry experts. While Udemy's market share is slightly lower than Coursera's, its revenue growth has been impressive, driven by a large user base and frequent course discounts. Udemy's key strategy for market expansion is to invest in marketing efforts to attract more learners and instructors to its platform.
Khan Academy, known for its free educational resources, has a significant market share in the non-profit sector of the EdTech industry. Despite its non-profit status, Khan Academy has seen steady revenue growth through donations and partnerships with foundations. Khan Academy's key strategy for market expansion is to continue offering high-quality educational content for free while seeking partnerships with schools and educational organizations.
In conclusion, the EdTech industry is thriving with opportunities for growth and innovation. By understanding the current market status, emerging trends, significant challenges, and major players in the industry, we can forecast a promising future for EdTech. However, it is important to address data privacy concerns, integrate EdTech platforms with traditional education systems, and focus on scalability to ensure sustained success. With the right strategies and a focus on continuous improvement and innovation, the EdTech industry can make a significant impact on the future of education.
As can be seen, the agent provides a highly detailed breakdown of the current state of the EdTech industry. However, as mentioned earlier, GPT 3.5-Turbo, which is the "brain" behind the free variant of AgentGPT, is somewhat limited compared to SOTA models. Let's now explore how an AI agent that uses SOTA models performs in comparison.
What Is Manus AI
Manus AI is an autonomous AI agent platform developed by the team at Manus Technologies. It is designed to enable users to create, customize, and deploy intelligent agents that can independently plan, reason, and execute tasks without constant oversight. The platform focuses on breaking down any problem, even the most complex, into manageable subtasks.
Manus AI has garnered attention for its performance, demonstrating its ability to handle highly complex tasks efficiently and with exceptional results. The platform has particularly attracted software engineers, as it proved the capability to autonomously create entire applications with minimal user input. Overall, Manus AI has proven to be an excellent platform for software developers. It significantly reduced the time required to design and create applications while generating high-quality code.
In addition to generating high-quality code, Manus AI can also navigate websites, use Vision models to scroll through pages, extract relevant data, and interact with necessary fields. If it runs into a captcha on a website that it needs, it will prompt the user to complete the verification, allowing the agent to continue its research.
When entering Manus, you will be directed to a page that looks like this:
Here, we can enter a prompt to assign Manus a task it needs to accomplish. Let's demonstrate by asking Manus to complete the following task:
I'd like a thorough analysis of Tesla stock, including:
Summary: Company overview, key metrics, performance data and investment recommendations
Financial Data: Revenue trends, profit margins, balance sheet and cash flow analysis
Market Sentiment: Analyst ratings, sentiment indicators and news impact
Technical Analysis: Price trends, technical indicators and support/resistance levels
Compare Assets: Market share and financial metrics vs. key competitors
Value Investor: Intrinsic value, growth potential and risk factors
Investment Thesis: SWOT analysis and recommendations for different investor types
Display the results of the analysis in a dashboard.
The result we receive is highly impressive. It is a dashboard with a detailed breakdown that looks like this:
To achieve this impressive result, Manus generated a comprehensive to-do list outlining all the steps required to complete the task we assigned to it.
After creating the list, Manus started executing various tasks. Throughout this process, numerous files are generated. In our case, it generated all of the following files while building the dashboard. The files are organized into:
- documents
- images
- code files
- links
These are the documents that were generated by Manus:
These are all images that were generated by Manus:
As can be seen, there are so many files that they don't all fit on the screen. Manus also provides us with access to all the code files it generated during the process.
Once again, there are so many files that they don't fit on the screen at once. Finally, to access the dashboard, we simply open the Links tab:
This problem-solving approach is what makes Manus AI both powerful and impressive.
AI agents are no longer exclusive to software developers. They are becoming increasingly accessible to the general public through various online platforms. In this article, we examined two such platforms. We began with AgentGPT, a user-friendly tool designed with simplicity in mind, making it ideal for everyday users. However, this ease of use comes with some limitations in advanced capabilities. In contrast, Manus AI offers significantly more power, capable of generating complete projects that include documentation, code, visuals, and even dashboards. These examples highlight a broader trend: we are gradually transitioning from the age of large language models to the era of AI agents.
