ChatGPT and Gemini are utilized to produce drafts, condense documents, analyze intricate subjects, or generate viral videos, while others are applying these models to tackle far more significant challenges. For instance, a study from the MIT Antibiotics-AI Project recently featured in the scientific journal Cell highlights not one but two AI methodologies that enabled scientists to identify unprecedented antibiotics that could potentially combat two perilous drug-resistant bacteria.
The new drugs were not developed solely by artificial intelligence models. Rather, the AI adhered to elaborate directives to identify molecules capable of eliminating Neisseria gonorrhoeae (gonorrhea) and Staphylococcus aureus (MRSA). These AI systems produced millions of potential chemical structures that could damage the bacteria and subsequently halt infections. In every instance, researchers employed specific criteria to refine the lists of compounds to suitable candidates. These criteria included stipulations that the resulting antibiotic should be non-harmful to humans and not possess similar characteristics to existing antibiotics that have ceased to be effective against the two bacteria. After imposing these requirements, the researchers identified a select number of promising candidates that exhibit potential in laboratory evaluations.
