Bookmark: OpenAI's ‘Deep Research’ Can Actually Make Professional Reports With Citations

A standout quote from the article is: According to OpenAI, “deep research was rated by domain experts to have automated multiple hours of difficult, manual investigation”> OpenAI’s ‘Deep Research’ Can Actually Make Professional Reports With Citations

OpenAI’s new feature, Deep Research, embedded within its ChatGPT, showcases the company’s advancements in autonomous AI functionalities. Targeting professionals within science, finance, engineering, and policy, this tool produces comprehensive reports akin to those crafted by research analysts. Operative through the o3 reasoning model, Deep Research can autonomously browse the internet, interpret and synthesize information from diverse digital sources, including text, images, and PDFs. Currently available for ChatGPT Pro users, it is slated to expand to other subscription levels. The service emphasizes high-level research, managing tasks that typically require multiple hours in a notably reduced timeframe. OpenAI highlights its utility in domains requiring detailed, technical insights, providing significant time savings in complex problem-solving scenarios. This innovative tool, however, has limitations, such as its potential to ‘hallucinate’ or inaccurately fabricate information, urging users to review its outputs critically. Despite these challenges, Deep Research has achieved high benchmarks in AI testing, outperforming other AI models in simulated assessments