With the rapid incorporation of artificial intelligence (AI) in scientific processes, research and development timelines are significantly shortened. Use cases range from drug invention to protein design and materials development, illustrating the profound potential of AI in reshaping the scientific landscape.
AI accelerates drug development
Artificial intelligence is proving to be a powerful tool in the field of drug discovery, particularly for neglected diseases. Graduate student Susana Vazquez-Torres at the University of Washington, for example, has begun using AI to develop new drugs for conditions such as snake bites. The process, once time consuming and laborious, has seen a dramatic acceleration thanks to AI. Vazquez-Torres, who only started her current project in February, already has several candidate drugs lined up.
The University of Washington is spearheading an innovative use of AI in the field of protein design. Here, around 200 scientists are leveraging AI to streamline the creation of new protein-based therapies. By using a form of AI known as diffusion modeling, they're able to efficiently create complex protein structures, skipping the formerly painstaking process of testing thousands of potential designs.
Despite the promising results in drug discovery and protein design, the application of AI in different scientific fields may be uneven. The use of AI in the development of new materials for renewable energy, for instance, is currently hindered by the lack of sufficient and organized data. Additionally, the multi-scale interactions involved in the properties of these materials add further complexity, making them more challenging to study using AI. Nonetheless, scientists like Maria Chan at Argonne National Laboratory maintain that AI can still provide valuable benefits in these fields.
AI as hypothesis hunters
Some researchers are envisioning AI taking a more central role in scientific discovery. Systems akin to ChatGPT are being developed with the goal of analyzing extensive scientific literature and generating fresh ideas or hypotheses. These AI systems could potentially uncover novel connections between studies and propose new lines of inquiry that researchers may not spot. This augments the potential for new scientific discoveries, provided that human researchers verify and follow up on these AI-generated hypotheses.
Envisioning AI-driven autonomous research
Looking ahead, some researchers, including Yolanda Gil at the University of Southern California, foresee an even more active role for AI in science. Gil visualizes automated systems capable of independently planning and executing experiments, which would involve developing new kinds of AI that are more adept at reasoning. These AI 'scientists' could continuously reassess data, offering real-time updates on research developments, and could also increase the reliability and systematic nature of scientific findings.
AI in science: An exciting tool, not a threat
Despite the rapid advancements in AI and its incursion into the scientific field, researchers like Vazquez-Torres consider it an exciting innovation rather than a threat to job security. AI is seen as a powerful tool enabling scientists to tackle a wider array of problems and create innovative solutions more easily and quickly. From new drug development to improving photosynthesis in plants and finding new compounds to degrade plastics, AI is opening up a world of possibilities for scientists.