On Sunday evening, OpenAI introduced a new mode for ChatGPT, called Deep Research, that allows the chatbot to conduct complex, multi-step research on the internet in real-time. This development marks a significant step towards OpenAI’s vision of building agentic AI capable of performing human-like tasks autonomously.
During a livestream announcement, OpenAI developers showcased the Deep Research mode, which is designed to give ChatGPT the ability to access data on the internet in real-time. The chatbot can now conduct thorough research with no latency constraints, although this comes at the cost of slightly longer response times.
The Deep Research mode is seen as a step closer to achieving artificial general intelligence (AGI), an AI capable of performing any intellectual task at the level of a human. OpenAI highlighted that this mode would be useful for individuals involved in knowledge-intensive work such as engineering, finance, literature, and biology.
The new mode is being offered to OpenAI Pro users at $200 per month, and the company plans to expand access to Plus tier users for $20 a month once safety requirements are met within the coming month.
While Deep Research allows ChatGPT to complete tasks ranging from 5 minutes to 30 minutes, depending on complexity, the mode demands high computing power. The system documents its outputs with clear citations and outlines the reasoning behind its responses. Despite the progress, OpenAI acknowledged that Deep Research is still prone to hallucinations, though at a lower rate compared to other ChatGPT models.
The release of Deep Research is a pivotal moment in the AI landscape. As OpenAI pushes toward AGI, this tool might be the first step in shifting AI from being a supportive tool to an autonomous decision-maker in various industries.
