GPT-5 is speeding up scientific research, but still can't be trusted to work alone, OpenAI warns

"OpenAI's recently released model, GPT-5 is showing promise in advancing scientific discovery. While user reactions to the new model in ChatGPT were less than stellar, it appears to be making more headway as a research assistant. In a new paper published Thursday, OpenAI detailed the ways GPT-5 "accelerated" research across a variety of case studies -- albeit with some limitations."

"It's the first report from OpenAI for Science, a team of internal researchers and recently-hired external academics that the company announced in September. The paper was also supported by researchers from several labs and universities, including Vanderbilt, UC Berkeley, Columbia, Cambridge, Oxford, The Jackson Laboratory, and others. According to a blog accompanying the paper, OpenAI for Science aims to help researchers save time by using frontier models to"

""Across these early studies, GPT-5 appears able to shorten parts of the research workflow when used by experts," the paper said. "It does not run projects or solve scientific problems autonomously, but it can expand the surface area of exploration and help researchers move faster toward correct results.""

GPT-5 assists researchers across disciplines by shortening parts of the research workflow when used by experts and expanding the surface area of exploration. The model can help researchers move faster toward correct results but cannot run projects or solve scientific problems autonomously. OpenAI compared GPT-5 with other models such as Claude and Gemini on real-world tasks. OpenAI for Science involved internal and external academic collaborators from institutions including Vanderbilt, UC Berkeley, Columbia, Cambridge, Oxford, and The Jackson Laboratory. Findings indicate GPT-5 is a useful research assistant rather than a replacement for human researchers. The results do not imply imminent AGI.

Read at ZDNET