Beyond the hype sparked by the mass deployment of generative AI after the rapid rise of ChatGPT, we need to reflect on how artificial intelligence is revolutionizing scientific research, with far more lasting effects on the future of humanity. Nobel laureates Demis Hassabis, John M. Jumper and David Baker have used AI to study proteins. David Baker developed software called Rosetta to achieve what was thought impossible: creating proteins that did not previously exist and may have completely new functions. Baker in 2003 published his discovery and released the code for his Rosetta software, which has since been used in new areas of science. Demis Hassabis and John Jumper, with their Google DeepMind team, later developed AlphaFold, an artificial intelligence model capable of accurately predicting the three-dimensional structure of proteins. The research group then created a protein, called Top7, has a unique structure that does not exist in nature. The design and creation of synthetic proteins is leading to major advances in fighting malaria, decomposing plastic waste, and could accelerate the discovery of new pharmaceutical solutions. Artificial intelligence can recreate natural compounds, but it can also imagine new materials. Among the major transformations taking place in the scientific world, the first concerns research productivity: AI allows large volumes of data to be analyzed very quickly, facilitating advances in complex fields such as medicine, astrophysics, chemistry, and psychology. For example, at the Heidelberg Institute, algorithms are being used to analyze galaxy data much faster than in the past, while at Stanford, AI is being used to address social problems, such as mapping poverty in Africa. The development of new sustainable technologies, which require extremely complex environmental impact analyses, could also accelerate thanks to artificial intelligence. AI is being used to identify patterns in masses of data too complex for humans to analyze and see. A perspective is conceivable in which researchers will no longer be the direct authors of our discoveries, but will be more like coaches training and leading a team of different artificial intelligences. Scientists “extended” by the artificial intelligence they themselves train.
Italiano