According to Greg Brockman on Twitter, GPT-5 Pro demonstrated the rapid potential of large language models (LLMs) in scientific discovery by suggesting the repurposing of a known drug to treat an otherwise untreatable food allergy within just 12 minutes. This recommendation matched the findings of a peer-reviewed scientific study that had not yet been published at the time, highlighting that advanced LLMs can independently arrive at validated scientific conclusions and dramatically accelerate the drug discovery process. As LLMs like GPT-5 Pro continue to improve, their application in pharmaceutical research opens up significant business opportunities, particularly for AI-driven drug discovery platforms and biotech firms seeking to shorten R&D timelines and reduce costs (source: Greg Brockman, x.com/DeryaTR_/status/1984083644437192737).

According to DeepMind's official announcement, users can now access a preview version of their new Science Model at deepmind.google.com/science/. This AI model is designed specifically to accelerate scientific research by enabling advanced data analysis and hypothesis generation, as detailed in their latest blog post (deepmind.google/discover/blog). The research paper (storage.googleapis.com/deepm) highlights the model's ability to process complex scientific datasets, making it a valuable tool for pharmaceutical development, materials discovery, and academic research. This launch marks a significant step towards practical AI deployment in scientific fields, offering business opportunities for AI integration in R&D workflows and expanding the market for AI-powered scientific solutions (Source: DeepMind Blog, 2024).