How GPT-5 helped immunologist Derya Unutmaz solve a 3-year-old mystery
Dr. Derya Unutmaz, an immunologist, leveraged GPT-5 Pro to unravel a three-year-old mystery concerning T-cell development and glucose. The AI model offered a crucial insight into how deoxyglucose impacts an important protein, accelerating Unutmaz's research into cancer and autoimmune diseases.
Immunologist Dr. Derya Unutmaz encountered a puzzling outcome in his 2022 experiment on T-cell development. He was investigating how glucose affects T cells, which are crucial for fighting cancer and infections. The results, particularly regarding the impact of deoxyglucose, were unclear to his lab at the time, leading them to set the experiment aside.
Years later, with the advent of GPT-5 Pro, Unutmaz revisited this unresolved mystery. He uploaded his experimental data to the AI model, seeking its analytical capabilities. GPT-5 Pro identified a key insight: deoxyglucose interferes with the construction of the IL-2 protein, which prevents T cells from becoming inflammatory Th17 cells. This explanation perfectly clarified the previous ambiguous results.
Further demonstrating its utility, GPT-5 Pro accurately predicted the outcome of an unpublished experiment on lymphoma-targeting T cells. This predictive power, coupled with its ability to generate crucial insights, transformed Unutmaz's view of AI. He now considers AI models as indispensable collaborators, akin to an extension of his own scientific capabilities.
AI, Unutmaz believes, can significantly accelerate biological research by streamlining literature reviews, refining hypotheses, and simulating experiments to identify the most promising avenues. However, he emphasizes that human expertise remains vital to evaluate the significance and plausibility of AI-generated insights, ensuring responsible and effective scientific advancement.
Related articles
Evaluating LLM Usage for Efficient and Explainable Numerical and Classified Implicit Sentiment Analysis of Product Desirability
A new paper explores the use of large language models for efficient and understandable implicit sentiment analysis, focusing on product desirability. This research delves into numerical and classified sentiment analysis, offering insights into advanced applications of LLMs.
Helping build shared standards for advanced AI
OpenAI is actively contributing to the development of shared standards for advanced AI, particularly through its involvement with the Appia Foundation. This initiative aims to create a critical "trust layer" for AI, enabling standardized assessment criteria across the AI value chain and fostering international cooperation. By establishing interoperable practices and technical understanding, the effort seeks to ensure safe and confident advancement of AI technologies globally.
The AI world is getting ‘loopy’
AI models are taking a significant leap forward with the adoption of "agentic loops," where AI agents continuously prompt each other to improve code and solve complex problems. This approach, though potentially resource-intensive, promises to unlock new levels of autonomous problem-solving and efficiency in AI applications.