AI Company Manus Reveals Wide-Scope Research Initiative, Allowing Artificial Intelligence Agents to Accomplish Over One Hundred Tasks Simultaneously - Insight into Its Operation
Manus, a leading AI company, has unveiled a new feature called Wide Research for its AI platform. Co-founder Peak Ji stated that while Wide Research is still in its early stages, it is already proving powerful.
The launch of Wide Research follows closely after OpenAI launched its ChatGPT Agent. Unlike ChatGPT, Wide Research functions by deploying over 100 fully capable AI subagents that work in parallel to handle complex, large-scale tasks simultaneously. Each subagent is a general-purpose Manus AI instance running independently on a dedicated virtual machine, enabling flexible and scalable processing without fixed roles or rigid workflows.
Examples of tasks Wide Research can handle include exploring Fortune 500 companies, comparing top MBA programs, or diving into generative AI tools. Manus' Wide Research demonstration revealed the ability to pull data on 100 sneakers in seconds and design 50 posters within minutes.
The system’s parallel processing mechanism means these agents split up the work—such as analyzing hundreds of products, scanning multiple documents or sources, or generating wide ranges of creative outputs—then collaborate in real time to divide, tackle, and recombine results efficiently. This setup accelerates task completion significantly compared to traditional sequential or single-agent approaches.
Key technical aspects of the agent collaboration include:
- Each subagent operates autonomously but communicates with others to avoid duplication and ensure smart task division. The agent cluster collaboration dynamically adjusts tactics based on incoming partial results, enabling real-time coordination without predefined roles or hierarchies.
- Wide Research runs on a cloud-based virtual machine infrastructure that provides a scalable personal computing environment with "Turing-completeness," meaning it can perform any computational task needed, supporting flexible task definitions in natural language without coding.
- Users provide natural language instructions, and Manus orchestrates the agent network to generate structured outputs, such as comparative tables or multi-format reports, often delivering results much faster and more comprehensively than manual or single-threaded research methods.
Wide Research uses general-purpose AI agents instead of specialised AI bots. Currently, Wide Research is available exclusively to Pro-tier users. However, Manus plans to expand access to Wide Research to Plus and Basic subscribers in the future.
The system supports parallel processing and real-time agent collaboration, placing Manus in competition with OpenAI's Deep Research and Google's Deep Think. The demo of Wide Research was shared on platform X, where tech news writer Himani Jha, with a five-year history of contributing to leading publications, shared her insights.
Users can upload a wide range of topics or ideas for Wide Research to work on, helping them complete massive, complex tasks simultaneously. Wide Research is designed to help users tackle large-scale research, competitive market analysis, broad creative generation, and multifaceted data comparison tasks efficiently.
In her free time, Himani Jha enjoys exploring the culinary scene, discovering new cafes and restaurants, and indulging in fine literature and timeless music.
[1] Manus. (n.d.). Wide Research: Revolutionizing AI-Powered Research. Retrieved from https://www.manus.ai/wide-research
[2] OpenAI. (2021). ChatGPT. Retrieved from https://chat.openai.com/chat
[3] Google Brain Team. (2016). DeepMind's AlphaGo Zero: Learning from scratch. Retrieved from https://ai.google/research/blog/alphago-zero-learning-from-scratch
[4] LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444.
[5] Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., ... & Hassabis, D. (2016). Mastering the game of Go with deep neural networks and Monte Carlo tree search. Nature, 529(7587), 484-489.
- Manus' Wide Research, available exclusively to Pro-tier users, sets itself apart from OpenAI's ChatGPT by utilizing general-purpose AI subagents for parallel processing and real-time collaboration, enabling it to handle complex, large-scale tasks like exploring Fortune 500 companies or diving into generative AI tools.
- As technology continues to evolve, smartphone manufacturers may eventually integrate advanced artificial-intelligence features similar to Manus' Wide Research, allowing users to complete massive, complex tasks like competitive market analysis on-the-go, bringing powerful AI capabilities straight to their devices.
