OpenAI has introduced a new AI feature called ChatGPT Deep Research, designed to facilitate complex and detailed research tasks across various fields, including finance, science, and engineering. This upgraded version of ChatGPT allows users to conduct extensive inquiries, utilizing information from multiple websites and sources. Deep Research is currently accessible to Pro users, with plans for broader rollout later on. Users can input queries and attach files for analysis, but the output is text-based for now, with future updates planned to include images and data visualizations.
Initial tests of the underlying O3 “River” AI model showed a 26.6% accuracy on rigorous academic questions, outperforming other models like Gemini Thinking and GROK-2. However, OpenAI acknowledges that Deep Research may still produce inaccuracies and struggle with distinguishing credible information from rumors. Users are advised to substantiate AI-generated findings with additional research. Interestingly, Google announced a similar AI feature under the same name two months earlier.
Source link
