On Sunday, OpenAI launched a new AI tool called “Deep Research,” designed to perform multi-step research tasks by gathering and synthesizing information from various online sources. This tool is powered by an advanced version of OpenAI’s upcoming o3 model, specifically optimized for web browsing and data analysis.
With Deep Research, users provide a prompt, and ChatGPT uses the tool to find, analyze, and compile data from various sources, including text, images, and PDFs, to produce a detailed research report comparable to the work of a research analyst. OpenAI claims that the tool can accomplish in minutes what would typically take a human several hours.
However, OpenAI has noted that Deep Research is still in its early stages and has some limitations. For instance, it may struggle to differentiate between authoritative information and rumors. Additionally, it faces challenges in accurately conveying uncertainty, often failing to present the level of confidence required in certain cases.
Deep Research is now available via the web version of ChatGPT, with plans to roll it out to mobile and desktop apps in February. This launch follows OpenAI’s introduction of another AI tool in January, called “Operator,” which is designed to assist with a variety of tasks, such as creating to-do lists or helping with vacation planning.