Alán Aspuru-Guzik, a prominent chemist, computer scientist, and supply chain expert at the University of Toronto, one of the project’s key leaders, suggests considering the potential collaboration between a robot and a human scientist in a chemistry experiment. The visionary goal of Aspuru-Guzik is to elevate traditional laboratory automation, striving to create an “AI scientist” capable of conducting and troubleshooting experiments, as well as providing insights into their outcomes.
Aspuru-Guzik and his team crafted Organa to be remarkably versatile. Instead of merely executing a single task or component of an experiment like a conventional fixed automation system would, it can perform a multi-step experiment upon cue. The system can be equipped with visualization tools that track progress and offer insights to optimize the experimental process.
According to Milad Abolhasani, a chemical and materials engineer at North Carolina State University who was not involved in the project, this is an early example of how one can have a bidirectional dialogue with an AI assistant for a robotic chemistry lab.
Most automated laboratory tools aren’t easily customized or reprogrammed to meet the precise needs of chemists, notes Florian Shkurti, a computer scientist at the University of Toronto and co-leader of the project. While some chemists may choose to develop programming skills, it’s not a requirement for their work. With Organa’s advanced technology, researchers can effortlessly share their findings through clear and concise verbal descriptions. As scientists input the robotic with their experimental targets and setup, Organa’s Large Language Model (LLM) translates these natural-language instructions into χDL codes – a standard chemical description language used in molecular design. The algorithm simplifies complex coding processes by structuring tasks into logical sequences and clear objectives, providing a roadmap for seamless execution. In cases where ambiguous instructions or unexpected outcomes arise, this could potentially trigger a red flag, prompting scientists to investigate and address the issue.
