OpenAI Wet-Lab Blog Is Pretty Good (nikomc.com)

OpenAI recently published a blog highlighting how they utilized GPT-5 to optimize the Gibson Assembly, a widely-used method in molecular biology for stitching DNA molecules together. By partnering with the biosecurity startup Red Queen Bio, OpenAI aimed to enhance cloning efficiency—a measure of how many successful colonies grow from a fixed amount of DNA. The AI's automated experimentation led to a remarkable 79-fold increase in cloning efficiency compared to traditional methods, achieved through innovative tweaks like adding two additional enzymes and modifying the DNA insertion protocol. This development is significant for the AI/ML community as it demonstrates AI's potential to accelerate biotechnological advancements, particularly in optimizing established methods that are often perceived as mundane. While some critics argue that similar results could be obtained by human researchers, the project underscores the importance of AI in driving faster iterations and improving the reliability of biological experiments. The incorporation of iterative AI methods into such fundamental practices could pave the way for heightened productivity and innovation in biotech research, marking a substantial step in the integration of AI into experimental biology.