Carnegie Mellon University’s Gabe Gomes has developed an AI system, Coscientist, that can plan and execute complex chemical reactions based on simple text prompts. Using the language model behind the chatbot ChatGPT, Coscientist can browse the internet for information, outline necessary procedures, and write code for a robot to run the experiment. On its first try, it successfully performed a Nobel Prize-winning reaction with a 50% yield.
The system could potentially automate the design-make-test cycle in labs, increasing productivity. However, the team is also aware of potential misuse and is working to prevent it. This was reported in Chemical and Engineering News.
AI System Conducts Complex Chemical Experiments
A new artificial intelligence (AI) system, named Coscientist, has been developed that can plan and execute complex chemical reactions based on simple text prompts. The system, which is driven by the language model behind the chatbot ChatGPT, can direct a robotic liquid handler to carry out the planned experiments. This technology could potentially streamline communication between humans and AI systems, thereby accelerating scientific research.
Coscientist works by taking a prompt such as “perform multiple Suzuki reactions”, and then use the internet to learn about the reactions. It scours relevant literature and hardware documentation for information, and within minutes, outlines the procedures necessary to perform these reactions. The AI then writes a code, which a robot uses to run the experiment.
Coscientist’s Successful Experiments
On its first attempt, Coscientist was able to successfully perform the complex Nobel Prize-winning palladium-catalyzed cross-coupling reaction named after Akira Suzuki with a 50% yield. In addition to this, the AI system could accurately plan procedures to synthesize common pharmaceutical compounds such as aspirin and ibuprofen.
The Role of Coscientist in the Lab
Gabe Gomes, a chemist and chemical engineer at Carnegie Mellon University, explains the significance of Coscientist’s capabilities. He states, “We are converting bits to atoms,” referring to the process of taking a natural language prompt, the bits, and converting it into an actual chemical reaction. However, Gomes and his team are fully aware of the potential illicit use of Coscientist and are collaborating with other researchers and policymakers to prevent such misuse.
The Future of AI in Chemistry
Tiago Rodrigues, a medicinal chemist at the University of Lisbon, believes that Coscientist fills an important communication gap necessary to achieve the long-standing goal of self-driving labs. He suggests that AI chemists such as Coscientist and another recently developed system, ChemCrow, could enable the full automation of the design-make-test cycle. Rodrigues states, “This can have a tremendous impact in terms of productivity since researchers can dedicate their time to other tasks.”
The Potential Impact of AI on Scientific Research
The development of AI systems like Coscientist and ChemCrow could revolutionize the field of scientific research. By automating complex experimental procedures, these systems could significantly increase productivity in labs, allowing researchers to focus on other important tasks. Furthermore, the ability of these AI systems to learn from the internet and relevant literature could potentially lead to discoveries and advancements in various scientific fields.
The Ethical Considerations of AI in Chemistry
While the potential benefits of AI systems like Coscientist are significant, it is also important to consider the ethical implications of such technology. The potential for misuse of these systems is a concern that must be addressed. As such, researchers and policymakers are working together to develop measures to prevent the illicit use of these AI systems. This highlights the importance of considering ethical issues in the development and implementation of new technologies.
This was reported in Chemical and Engineering News.
