Artificial intelligence is taking over the world of science. Researchers in the US recently conducted an experiment with the help of a fully autonomous laboratory, piloted by an artificial intelligence model and robotic equipment. In addition to speeding up the scientific process, this new approach could lead to solutions that humans might not have considered on their own.

       An article in Nature details this fully automated laboratory experiment. It recounts how researchers at the University of Wisconsin-Madison used a machine-learning model to propose a whole host of protein sequence modifications, all of which could then be tested. In the end, this process created enzymes capable of operating at temperatures at least 12°C higher than those at the outset.

       In practice, this laboratory was made up of robotic equipment run by a relatively simple artificial intelligence model. Over time, the system designed different protein modifications, each more refined and each time tested by the robots. Right up to the end, this automated process optimised the results.

       This study process took almost six months to complete, with only occasional repairs requiring human intervention. “A human might spend up to a year doing the same work," the Nature article states.

       This type of autonomous, or “self-driving” laboratory paves the way for rapid and effective advances, at least in protein research. However, the idea of AI suggesting modifications and robots systematically testing them to optimize their effectiveness could well be extended to other types of research.

       The pharmaceutical industry, for its part, is already working in this field. A startup called Insilico Medicine has used artificial intelligence to design new drugs for rare diseases, more specifically a new compound to treat idiopathic pulmonary fibrosis (IPF). In this particular case, AI enabled researchers to save an enormous amount of time, given that the search for drugs for this type of disease is usually extremely time-consuming and costly for laboratories.

       In a completely different field, but based on the same principle, artificial intelligence recently made it possible to develop a more lithium-efficient material that could one day equip a whole new generation of batteries. In this case, artificial intelligence considerably reduced research time by managing to study 32 million potential candidates in just 80 hours. – AFP Relaxnews

       StarPicks

       Maximise your EPF savings with voluntary contributions